id
stringlengths 9
16
| submitter
stringlengths 1
64
⌀ | authors
stringlengths 3
62.8k
| title
stringlengths 1
382
| comments
stringlengths 1
1.15k
⌀ | journal-ref
stringlengths 1
557
⌀ | doi
stringlengths 7
153
⌀ | report-no
stringlengths 1
509
⌀ | categories
stringlengths 5
125
| license
stringclasses 9
values | abstract
stringlengths 6
6.09k
| versions
listlengths 1
187
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-04-25 00:00:00
| authors_parsed
sequencelengths 1
2.83k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2504.08254 | Emiliano De Cristofaro | Georgi Ganev, Meenatchi Sundaram Muthu Selva Annamalai, Sofiane
Mahiou, Emiliano De Cristofaro | Understanding the Impact of Data Domain Extraction on Synthetic Data
Privacy | Accepted to the Synthetic Data x Data Access Problem workshop
(SynthData), part of ICLR 2025 | null | null | null | cs.CR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Privacy attacks, particularly membership inference attacks (MIAs), are widely
used to assess the privacy of generative models for tabular synthetic data,
including those with Differential Privacy (DP) guarantees. These attacks often
exploit outliers, which are especially vulnerable due to their position at the
boundaries of the data domain (e.g., at the minimum and maximum values).
However, the role of data domain extraction in generative models and its impact
on privacy attacks have been overlooked. In this paper, we examine three
strategies for defining the data domain: assuming it is externally provided
(ideally from public data), extracting it directly from the input data, and
extracting it with DP mechanisms. While common in popular implementations and
libraries, we show that the second approach breaks end-to-end DP guarantees and
leaves models vulnerable. While using a provided domain (if representative) is
preferable, extracting it with DP can also defend against popular MIAs, even at
high privacy budgets.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 04:35:24 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 02:34:24 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ganev",
"Georgi",
""
],
[
"Annamalai",
"Meenatchi Sundaram Muthu Selva",
""
],
[
"Mahiou",
"Sofiane",
""
],
[
"De Cristofaro",
"Emiliano",
""
]
] |
2504.08256 | Shiyi Ding | Shiyi Ding, Ying Chen | RAG-VR: Leveraging Retrieval-Augmented Generation for 3D Question
Answering in VR Environments | GenAI-XR 2025 Workshop, co-located with 2025 IEEE Conference on
Virtual Reality and 3D User Interfaces (VR) | null | null | null | cs.IR cs.AI cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in large language models (LLMs) provide new opportunities for
context understanding in virtual reality (VR). However, VR contexts are often
highly localized and personalized, limiting the effectiveness of
general-purpose LLMs. To address this challenge, we present RAG-VR, the first
3D question-answering system for VR that incorporates retrieval-augmented
generation (RAG), which augments an LLM with external knowledge retrieved from
a localized knowledge database to improve the answer quality. RAG-VR includes a
pipeline for extracting comprehensive knowledge about virtual environments and
user conditions for accurate answer generation. To ensure efficient retrieval,
RAG-VR offloads the retrieval process to a nearby edge server and uses only
essential information during retrieval. Moreover, we train the retriever to
effectively distinguish among relevant, irrelevant, and hard-to-differentiate
information in relation to questions. RAG-VR improves answer accuracy by
17.9%-41.8% and reduces end-to-end latency by 34.5%-47.3% compared with two
baseline systems.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 04:55:50 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 01:31:40 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ding",
"Shiyi",
""
],
[
"Chen",
"Ying",
""
]
] |
2504.08295 | Sagar Mandal | Sagar Mandal | Exploring the Relationships Between the Divisors of Friends of $10$ | 10 pages | News Bulletin of Calcutta Mathematical Society 48 (1-3) (2025),
21-32 | 10.5281/zenodo.15206287 | null | math.GM | http://creativecommons.org/licenses/by/4.0/ | A solitary number is a positive integer that shares its abundancy index only
with itself. $10$ is the smallest positive integer suspected to be solitary,
but no proof has been established so far. In this paper, we prove that not all
half of the exponents of the prime divisors of a friend of 10 are congruent to
$1$ modulo $3$. Furthermore, we prove that if $F=5^{2a}\cdot Q^2$ ($Q$ is an
odd positive integer coprime to $15$) is a friend of $10$, then
$\sigma(5^{2a})+\sigma(Q^2)$ is congruent to $6$ modulo $8$ if and only if $a$
is even, and $\sigma(5^{2a}) + \sigma(Q^2)$ is congruent to $2$ modulo $8$ if
and only if $a$ is odd. In addition, if we set $Q={\displaystyle
\prod_{i=2}^{\omega(F)}}p_{i}^{a_i}$ and $a=a_1$, where $p_i$ are prime
numbers, then we establish that
$$F>\frac{25}{81}\cdot\prod_{i=1}^{\omega(F)}(2a_i + 1)^2,$$ in particular $F>
625\cdot 9^{\omega(F)-3}.$
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 06:52:45 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mandal",
"Sagar",
""
]
] |
2504.08327 | Zden\v{e}k Dvo\v{r}\'ak | Zden\v{e}k Dvo\v{r}\'ak and Bernard Lidick\'y and Bojan Mohar | On a conjecture concerning 4-coloring of graphs with one crossing | 51 pages, 6 figures Metadata update (fixing a typo in the abstract) | null | null | null | math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We conjecture that every graph of minimum degree five with no separating
triangles and drawn in the plane with one crossing is 4-colorable. In this
paper, we use computer enumeration to show that this conjecture holds for all
graphs with at most 28 vertices, explore the consequences of this conjecture
and provide some insights on how it could be proved.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 07:50:40 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 07:50:23 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Dvořák",
"Zdeněk",
""
],
[
"Lidický",
"Bernard",
""
],
[
"Mohar",
"Bojan",
""
]
] |
2504.08338 | Zhaopeng Zhang | Zhaopeng Zhang, Shizhen Wu, Chenfeng Guo, Yongchun Fang, Jianda Han,
and Xiao Liang | RINGO: Real-time Navigation with a Guiding Trajectory for Aerial
Manipulators in Unknown Environments | 9 pages, 15 figures | null | null | null | cs.RO | http://creativecommons.org/licenses/by/4.0/ | Motion planning for aerial manipulators in constrained environments has
typically been limited to known environments or simplified to that of
multi-rotors, which leads to poor adaptability and overly conservative
trajectories. This paper presents RINGO: Real-time Navigation with a Guiding
Trajectory, a novel planning framework that enables aerial manipulators to
navigate unknown environments in real time. The proposed method simultaneously
considers the positions of both the multi-rotor and the end-effector. A
pre-obtained multi-rotor trajectory serves as a guiding reference, allowing the
end-effector to generate a smooth, collision-free, and workspace-compatible
trajectory. Leveraging the convex hull property of B-spline curves, we
theoretically guarantee that the trajectory remains within the reachable
workspace. To the best of our knowledge, this is the first work that enables
real-time navigation of aerial manipulators in unknown environments. The
simulation and experimental results show the effectiveness of the proposed
method. The proposed method generates less conservative trajectories than
approaches that consider only the multi-rotor.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 08:07:58 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 16:23:12 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Zhaopeng",
""
],
[
"Wu",
"Shizhen",
""
],
[
"Guo",
"Chenfeng",
""
],
[
"Fang",
"Yongchun",
""
],
[
"Han",
"Jianda",
""
],
[
"Liang",
"Xiao",
""
]
] |
2504.08410 | Mingzhi Pei | Mingzhi Pei, Xu Cao, Xiangyi Wang, Heng Guo, Zhanyu Ma | PMNI: Pose-free Multi-view Normal Integration for Reflective and
Textureless Surface Reconstruction | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reflective and textureless surfaces remain a challenge in multi-view 3D
reconstruction. Both camera pose calibration and shape reconstruction often
fail due to insufficient or unreliable cross-view visual features. To address
these issues, we present PMNI (Pose-free Multi-view Normal Integration), a
neural surface reconstruction method that incorporates rich geometric
information by leveraging surface normal maps instead of RGB images. By
enforcing geometric constraints from surface normals and multi-view shape
consistency within a neural signed distance function (SDF) optimization
framework, PMNI simultaneously recovers accurate camera poses and high-fidelity
surface geometry. Experimental results on synthetic and real-world datasets
show that our method achieves state-of-the-art performance in the
reconstruction of reflective surfaces, even without reliable initial camera
poses.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 10:16:55 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 01:50:54 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Pei",
"Mingzhi",
""
],
[
"Cao",
"Xu",
""
],
[
"Wang",
"Xiangyi",
""
],
[
"Guo",
"Heng",
""
],
[
"Ma",
"Zhanyu",
""
]
] |
2504.08424 | Iuliia Grotova | Iuliia Grotova, Arne Rau, Pietro Baldini, Adelle J. Goodwin, Zhu Liu,
Andrea Merloni, Mara Salvato, Gemma E. Anderson, Riccardo Arcodia, Johannes
Buchner, Mirko Krumpe, Adam Malyali, Megan Masterson, James C. A.
Miller-Jones, Kirpal Nandra, Raphael Shirley | The population of tidal disruption events discovered with eROSITA | 23 pages, 12 figures. Accepted for publication in A&A. Added a
reference | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | This paper presents a systematic study of X-ray-selected canonical tidal
disruption events (TDEs) discovered in the western Galactic hemisphere of the
first two eROSITA all-sky surveys (eRASS1 and eRASS2) performed between Dec
2019 and Dec 2020. We compiled a TDE sample from the catalog of eROSITA's
extragalactic transients and variables eRO-ExTra, which includes X-ray sources
with a variability significance and fractional amplitude over four between
eRASS1 and eRASS2, not associated with known AGNs. Each X-ray source is
associated with an optical counterpart from the Legacy Survey DR10. Canonical
TDEs were selected based on their X-ray light-curve properties (single flare or
decline), soft X-ray spectra ($\Gamma>3$), and the absence of archival X-ray
variability and AGN signatures in their host photometry and spectroscopy. The
sample includes 31 X-ray-selected TDE candidates with redshifts of $0.02<
z<0.34$ and luminosities of $5.7 \times 10^{41}<L_X<5.3 \times 10^{44}$ erg/s
in the 0.2-6.0 keV rest frame, of which 30 are canonical TDEs and one is an
off-nuclear TDE candidate. The derived X-ray luminosity function is best fit by
a double power law with a luminosity break at $10^{44}$ erg/s, corresponding to
the Eddington-limiting prediction. This corresponds to a TDE volumetric rate of
$ (2.3^{+1.2}_{-0.9})\times10^{-7}\,Mpc^{-3} yr^{-1}$ ($\approx1.2\times
10^{-5}$ events per galaxy per year). TDE host galaxies show a green-valley
overdensity. In addition, 20%, 30%, and 15% of the sample exhibit flares in the
optical, mid-infrared (mid-IR), or radio bands, respectively. We discuss the
differences between X-ray, optical, and mid-IR TDE populations and the origins
of multiwavelength flares in the context of the obscuring envelope and
stream-stream collision models. Finally, we highlight TDE subpopulations that
are not included in the canonical sample and should be explored in the future.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 10:30:18 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 08:03:24 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Grotova",
"Iuliia",
""
],
[
"Rau",
"Arne",
""
],
[
"Baldini",
"Pietro",
""
],
[
"Goodwin",
"Adelle J.",
""
],
[
"Liu",
"Zhu",
""
],
[
"Merloni",
"Andrea",
""
],
[
"Salvato",
"Mara",
""
],
[
"Anderson",
"Gemma E.",
""
],
[
"Arcodia",
"Riccardo",
""
],
[
"Buchner",
"Johannes",
""
],
[
"Krumpe",
"Mirko",
""
],
[
"Malyali",
"Adam",
""
],
[
"Masterson",
"Megan",
""
],
[
"Miller-Jones",
"James C. A.",
""
],
[
"Nandra",
"Kirpal",
""
],
[
"Shirley",
"Raphael",
""
]
] |
2504.08426 | Adelle Goodwin Dr | A. J. Goodwin, M. Burn, G. E. Anderson, J. C. A. Miller-Jones, I.
Grotova, P. Baldini, Z. Liu, A. Malyali, A. Rau, M. Salvato | A systematic analysis of the radio properties of 22 X-ray selected tidal
disruption event candidates with the Australia Telescope Compact Array | 23 pages, accepted for publication in ApJS | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | We present a systematic analysis of the radio properties of an X-ray selected
sample of tidal disruption event (TDE) candidates discovered by the eROSITA
telescope. We find radio sources coincident with half of the transient events
(11 TDEs), with 8 radio sources showing statistically significant variability
over a 6-month period. We model the radio spectra of 6 sources with
sufficiently bright radio emission and find the sources show radio spectra
consistent with optically thin synchrotron emission and radio outflow minimum
radii of $10^{16}$--$10^{17}$ cm, velocities 0.01--0.05 c, and energies
$10^{48}$--$10^{51}$ erg. On comparison with the radio properties of an
optically-selected TDE sample at similar late times, we find no significant
difference in the radio luminosity range or radio detection rate. We find a
tentative positive trend with peak radio and X-ray luminosity, but require
further observations to determine if this is real or due to observational bias
due to the large range in distances of the events. Interestingly, none of the
X-ray selected events show late rising radio emission, compared to 45% of
radio-detected sources of an optically-selected sample that showed late rising
radio emission. We propose that this may indicate that many TDEs launch radio
outflows at or near peak X-ray luminosity, which can be significantly delayed
from peak optical luminosity. This study presents the first systematic analysis
of the radio properties of an X-ray selected sample of TDEs, and gives insight
into the possible link between the physical processes that power X-ray and
radio emission in TDEs.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 10:34:16 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 08:26:26 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Goodwin",
"A. J.",
""
],
[
"Burn",
"M.",
""
],
[
"Anderson",
"G. E.",
""
],
[
"Miller-Jones",
"J. C. A.",
""
],
[
"Grotova",
"I.",
""
],
[
"Baldini",
"P.",
""
],
[
"Liu",
"Z.",
""
],
[
"Malyali",
"A.",
""
],
[
"Rau",
"A.",
""
],
[
"Salvato",
"M.",
""
]
] |
2504.08507 | The CMS Collaboration | CMS Collaboration | Search for jet quenching with dijets from high-multiplicity pPb
collisions at $\sqrt{s_\mathrm{NN}}$ = 8.16 TeV | Submitted to the Journal of High Energy Physics. All figures and
tables can be found at
http://cms-results.web.cern.ch/cms-results/public-results/publications/HIN-23-010
(CMS Public Pages) | null | null | CMS-HIN-23-010, CERN-EP-2025-043 | nucl-ex hep-ex | http://creativecommons.org/licenses/by/4.0/ | The first measurement of the dijet transverse momentum balance $x_j$ in
proton-lead (pPb) collisions at a nucleon-nucleon center-of-mass energy of
$\sqrt{s_\mathrm{NN}}$ = 8.16 TeV is presented. The $x_j$ observable, defined
as the ratio of the subleading over leading jet transverse momentum in a dijet
pair, is used to search for jet quenching effects. The data, corresponding to
an integrated luminosity of 174.6 nb$^{-1}$, were collected with the CMS
detector in 2016. The $x_j$ distributions and their average values are studied
as functions of the charged-particle multiplicity of the events and for various
dijet rapidity selections. The latter enables probing hard scattering of
partons carrying distinct nucleon momentum fractions \x in the proton- and
lead-going directions. The former, aided by the high-multiplicity triggers,
allows probing for potential jet quenching effects in high-multiplicity events
(with up to 400 charged particles), for which collective phenomena consistent
with quark-gluon plasma (QGP) droplet formation were previously observed. The
ratios of $x_j$ distributions for high- to low-multiplicity events are used to
quantify the possible medium effects. These ratios are consistent with
simulations of the hard-scattering process that do not include QGP production.
These measurements set an upper limit on medium-induced energy loss of the
subleading jet of 1.26% of its transverse momentum at the 90% confidence level
in high multiplicity pPb events.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 13:19:20 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"CMS Collaboration",
"",
""
]
] |
2504.08535 | Yankai Lin | Yankai Lin, Michelle S. Chong, Carlos Murguia | Secondary Safety Control for Systems with Sector Bounded Nonlinearities
[Extended Version] | Supplementary material for the Automatica submission | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | We consider the problem of safety verification and safety-aware controller
synthesis for systems with sector bounded nonlinearities. We aim to keep the
states of the system within a given safe set under potential actuator and
sensor attacks. Specifically, we adopt the setup that a controller has already
been designed to stabilize the plant. Using invariant sets and barrier
certificate theory, we first give sufficient conditions to verify the safety of
the closed-loop system under attacks. Furthermore, by using a subset of sensors
that are assumed to be free of attacks, we provide a synthesis method for a
secondary controller that enhances the safety of the system. The sufficient
conditions to verify safety are derived using Lyapunov-based tools and the
S-procedure. Using the projection lemma, the conditions are then formulated as
linear matrix inequality (LMI) problems which can be solved efficiently.
Lastly, our theoretical results are illustrated through numerical simulations.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 13:44:22 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 17:38:53 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lin",
"Yankai",
""
],
[
"Chong",
"Michelle S.",
""
],
[
"Murguia",
"Carlos",
""
]
] |
2504.08571 | Taito Shimoji | Taito Shimoji | Morgan's mixed Hodge structures on $p$-filiform Lie algebras and
low-dimensional nilpotent Lie algebras | 17pages and 4 tables. This work is scheduled to be presented at "New
Developments of Transformation Groups" (RIMS). Comments welcome! | null | null | null | math.DG math.AG math.AT math.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The aim of this paper is to show that the fundamental group of any smooth
complex algebraic variety cannot be realized as a lattice of any simply
connected nilpotent Lie group whose Lie algebra is $p$-filiform Lie algebra
such that neither abelian nor $2$-step nilpotent. Moreover, we provide a
sufficient condition for a lattice in a simply connected nilpotent Lie group of
dimension up to $6$ not to be isomorphic to the fundamental group of any smooth
complex algebraic variety.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 14:26:43 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 01:34:33 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Shimoji",
"Taito",
""
]
] |
2504.08575 | Martin Zimmermann | Sarah Winter and Martin Zimmermann | Prophecies all the Way: Game-based Model-Checking for HyperQPTL beyond
$\forall^*\exists^*$ | null | null | null | null | cs.LO cs.FL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Model-checking HyperLTL, a temporal logic expressing properties of sets of
traces with applications to information-flow based security and privacy, has a
decidable, but TOWER-complete, model-checking problem. In the classical
algorithm, the complexity manifests itself with a need for the complementation
of automata over infinite words. To overcome this aforementioned need, a
game-based alternative for the $\forall^*\exists^*$-fragment was recently
presented.
Here, we employ imperfect information-games to extend the game-based approach
to full HyperQPTL, i.e., we allow arbitrary quantifier prefixes and
quantification over propositions, which allows us to express every
$\omega$-regular hyperproperty. As a byproduct of our game-based algorithm, we
obtain finite-state implementations of Skolem functions via transducers with
lookahead that explain satisfaction or violation of HyperQPTL properties.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 14:29:00 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 09:48:06 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Winter",
"Sarah",
""
],
[
"Zimmermann",
"Martin",
""
]
] |
2504.08589 | Alexander Zupan | Ben Clingenpeel, Zongzheng Dai, Gabriel Diraviam, Kareem Jaber,
Krishnendu Kar, Ziyun Liu, Teo Miklethun, Haritha Nagampoozhy, Michael Perry,
Moses Samuelson-Lynn, Eli Seamans, Ana Wright, Nicole Xie, Ruiqi Zou,
Alexander Zupan | Colorings of symmetric unions and partial knots | 13 pages, 7 figures, comments welcome! | null | null | null | math.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Motivated by work of Kinoshita and Teraska, Lamm introduced the notion of a
symmetric union, which can be constructed from a partial knot $J$ by
introducing additional crossings to a diagram of $J \# -\!J$ along its axis of
symmetry. If both $J$ and $J'$ are partial knots for different symmetric union
presentations of the same ribbon knot $K$, the knots $J$ and $J'$ are said to
be symmetrically related. Lamm proved that if $J$ and $J'$ are symmetrically
related, then $\det J = \det J'$, asking whether the converse is true. In this
article, we give a negative answer to Lamm's question, constructing for any
natural number $m$ a family of $2^m$ knots with the same determinant but such
that no two knots in the family are symmetrically related. This result is a
corollary to our main theorem, that if $J$ is the partial knot in a symmetric
union presentation for $K$, then $\text{col}_p(J) \leq \text{col}_p(K) \leq
\frac{(\text{col}_p(J))^2}{2}$, where $\text{col}_p(\cdot )$ denotes the number
of $p$-colorings of a knot.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 14:46:53 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Clingenpeel",
"Ben",
""
],
[
"Dai",
"Zongzheng",
""
],
[
"Diraviam",
"Gabriel",
""
],
[
"Jaber",
"Kareem",
""
],
[
"Kar",
"Krishnendu",
""
],
[
"Liu",
"Ziyun",
""
],
[
"Miklethun",
"Teo",
""
],
[
"Nagampoozhy",
"Haritha",
""
],
[
"Perry",
"Michael",
""
],
[
"Samuelson-Lynn",
"Moses",
""
],
[
"Seamans",
"Eli",
""
],
[
"Wright",
"Ana",
""
],
[
"Xie",
"Nicole",
""
],
[
"Zou",
"Ruiqi",
""
],
[
"Zupan",
"Alexander",
""
]
] |
2504.08593 | Jian He Low | Low Jian He, Harry Walsh, Ozge Mercanoglu Sincan, Richard Bowden | Hands-On: Segmenting Individual Signs from Continuous Sequences | Accepted in the 19th IEEE International Conference on Automatic Face
and Gesture Recognition | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This work tackles the challenge of continuous sign language segmentation, a
key task with huge implications for sign language translation and data
annotation. We propose a transformer-based architecture that models the
temporal dynamics of signing and frames segmentation as a sequence labeling
problem using the Begin-In-Out (BIO) tagging scheme. Our method leverages the
HaMeR hand features, and is complemented with 3D Angles. Extensive experiments
show that our model achieves state-of-the-art results on the DGS Corpus, while
our features surpass prior benchmarks on BSLCorpus.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 14:52:59 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 08:07:48 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"He",
"Low Jian",
""
],
[
"Walsh",
"Harry",
""
],
[
"Sincan",
"Ozge Mercanoglu",
""
],
[
"Bowden",
"Richard",
""
]
] |
2504.08597 | Valentin Leeb | Peru d'Ornellas, Valentin Leeb, Adolfo G. Grushin, Johannes Knolle | Altermagnetism Without Crystal Symmetry | null | null | null | null | cond-mat.str-el cond-mat.dis-nn | http://creativecommons.org/licenses/by/4.0/ | Altermagnetism is a collinear magnetic order in which opposite spin species
are exchanged under a real-space rotation. Hence, the search for physical
realizations has focussed on crystalline solids with specific rotational
symmetry. Here, we show that altermagnetism can also emerge in non-crystalline
systems, such as amorphous solids, despite the lack of global rotational
symmetries. We construct a Hamiltonian with two directional orbitals per site
on an amorphous lattice with interactions that are invariant under spin
rotation. Altermagnetism then arises due to spontaneous symmetry breaking in
the spin and orbital degrees of freedom around each atom, displaying a common
point group symmetry. This form of altermagnetism exhibits anisotropic spin
transport and spin spectral functions, both experimentally measurable. Our
mechanism generalizes to any lattice and any altermagnetic order, opening the
search for altermagnetic phenomena to non-crystalline systems.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 14:55:34 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 08:04:59 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"d'Ornellas",
"Peru",
""
],
[
"Leeb",
"Valentin",
""
],
[
"Grushin",
"Adolfo G.",
""
],
[
"Knolle",
"Johannes",
""
]
] |
2504.08598 | Toonyawat Angkhanawin | Toonyawat Angkhanawin, Aydin Deger, Jonathan D. Pritchard, C. Stuart
Adams | Graph Coloring via Quantum Optimization on a Rydberg-Qudit Atom Array | 18 pages, 10 figures | null | null | null | quant-ph physics.atom-ph | http://creativecommons.org/licenses/by/4.0/ | Neutral atom arrays have emerged as a versatile candidate for the embedding
of hard classical optimization problems. Prior work has focused on mapping
problems onto finding the maximum independent set of weighted or unweighted
unit disk graphs. In this paper we introduce a new approach to solving
natively-embedded vertex graph coloring problems by performing coherent
annealing with Rydberg-qudit atoms, where different same-parity Rydberg levels
represent a distinct label or color. We demonstrate the ability to robustly
find optimal graph colorings for chromatic numbers up to the number of distinct
Rydberg states used, in our case $k=3$. We analyze the impact of both the
long-range potential tails and residual inter-state interactions, proposing
encoding strategies that suppress errors in the resulting ground states. We
discuss the experimental feasibility of this approach and propose extensions to
solve higher chromatic number problems, providing a route towards direct
solution of a wide range of real-world integer optimization problems using
near-term neutral atom hardware.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 14:55:35 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Angkhanawin",
"Toonyawat",
""
],
[
"Deger",
"Aydin",
""
],
[
"Pritchard",
"Jonathan D.",
""
],
[
"Adams",
"C. Stuart",
""
]
] |
2504.08625 | Maciej Matys | Maciej Matys, Atsushi Yamada, Yoichi Kamada, and Toshihiro Ohki | A simple method for detection and quantitative estimation of deep levels
in a barrier layer of AlGaN/GaN HEMT structures by analysis of light induced
threshold voltage shift | Fixed the typo in the title | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | The characterization of deep levels in AlGaN/GaN heterostructures is one of
the most important problems in GaN high electron mobility transistors (HEMTs)
technology. This work reports on a technique for determination of deep level
concentration in AlGaN/GaN HEMT structures. The proposed method is relatively
simple, and it is based on the detection of free holes created by optically
induced transitions of electrons from the deep levels to the conduction band.
The developed method can detect and provide quantitative estimation of deep
level traps in a barrier layer of AlGaN/GaN HEMT structures. Furthermore, it
provides a framework for analysis of light induced threshold voltage shift,
which includes an important experimental criterion of determination whether the
holes are generated or not in the AlGaN/GaN HEMT structures by sub-band gap
illumination. The method was verified by applications it to a study of the deep
levels in GaN HEMTs grown on various substrates, i.e. SiC and GaN.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 15:26:39 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 03:46:05 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Matys",
"Maciej",
""
],
[
"Yamada",
"Atsushi",
""
],
[
"Kamada",
"Yoichi",
""
],
[
"Ohki",
"Toshihiro",
""
]
] |
2504.08634 | Niccol\`o Cribiori | David Andriot, Niccol\`o Cribiori, Thomas Van Riet | Scale separation, rolling solutions and entropy bounds | 15 pages | null | null | null | hep-th astro-ph.CO gr-qc hep-ph | http://creativecommons.org/licenses/by/4.0/ | We revisit scale separation for compactifications of ten- and
eleven-dimensional supergravity. For cosmological solutions rolling down
flux-generated potentials, we observe that scale separation is achieved as time
flows, and is fairly generic. This is realized without the need of orientifolds
nor corrections to the classical supergravity approximation. We then confront
scale separation with the Covariant Entropy Bound (CEB) and the CKN bound. We
show that a naive application of these bounds to vacua hints at the existence
of at least two extra dimensions. For rolling solutions, we observe that the
CEB is not always respected, but since these examples lack a cosmic horizon,
the application of entropy bounds remains delicate.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 15:36:32 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Andriot",
"David",
""
],
[
"Cribiori",
"Niccolò",
""
],
[
"Van Riet",
"Thomas",
""
]
] |
2504.08696 | Timothy Bula | Timothy Bula, Saurabh Pujar, Luca Buratti, Mihaela Bornea, Avirup Sil | SeaView: Software Engineering Agent Visual Interface for Enhanced
Workflow | 8 pages, 5 figures | null | null | null | cs.SE cs.LG | http://creativecommons.org/licenses/by/4.0/ | Auto-regressive LLM-based software engineering (SWE) agents, henceforth SWE
agents, have made tremendous progress (>60% on SWE-Bench Verified) on
real-world coding challenges including GitHub issue resolution. SWE agents use
a combination of reasoning, environment interaction and self-reflection to
resolve issues thereby generating "trajectories". Analysis of SWE agent
trajectories is difficult, not only as they exceed LLM sequence length
(sometimes, greater than 128k) but also because it involves a relatively
prolonged interaction between an LLM and the environment managed by the agent.
In case of an agent error, it can be hard to decipher, locate and understand
its scope. Similarly, it can be hard to track improvements or regression over
multiple runs or experiments. While a lot of research has gone into making
these SWE agents reach state-of-the-art, much less focus has been put into
creating tools to help analyze and visualize agent output. We propose a novel
tool called SeaView: Software Engineering Agent Visual Interface for Enhanced
Workflow, with a vision to assist SWE-agent researchers to visualize and
inspect their experiments. SeaView's novel mechanisms help compare experimental
runs with varying hyper-parameters or LLMs, and quickly get an understanding of
LLM or environment related problems. Based on our user study, experienced
researchers spend between 10 and 30 minutes to gather the information provided
by SeaView, while researchers with little experience can spend between 30
minutes to 1 hour to diagnose their experiment.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 17:03:58 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 16:02:38 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Bula",
"Timothy",
""
],
[
"Pujar",
"Saurabh",
""
],
[
"Buratti",
"Luca",
""
],
[
"Bornea",
"Mihaela",
""
],
[
"Sil",
"Avirup",
""
]
] |
2504.08698 | Hossein Shojaei | Hossein Shojaei, Hamid Rahmanei, Seyed Hossein Sadati | Performance Evaluation of Trajectory Tracking Controllers for a
Quadruped Robot Leg | Published in IEEE Xplore | null | 10.1109/ICRoM60803.2023.10412397 | null | cs.RO cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The complexities in the dynamic model of the legged robots make it necessary
to utilize model-free controllers in the task of trajectory tracking. In This
paper, an adaptive transpose Jacobian approach is proposed to deal with the
dynamic model complexity, which utilizes an adaptive PI-algorithm to adjust the
control gains. The performance of the proposed control algorithm is compared
with the conventional transpose Jacobian and sliding mode control algorithms
and evaluated by the root mean square of the errors and control input energy
criteria. In order to appraise the effectiveness of the proposed control
system, simulations are carried out in MATLAB/Simulink software for a quadruped
robot leg for semi-elliptical path tracking. The obtained results show that the
proposed adaptive transpose Jacobian reduces the overshoot and root mean square
of the errors and at the same time, decreases the control input energy.
Moreover, transpose Jacobin and adaptive transpose Jacobian are more robust to
changes in initial conditions compared to the conventional sliding mode
control. Furthermore, sliding mode control performs well up to 20%
uncertainties in the parameters due to its model-based nature, whereas the
transpose Jacobin and the proposed adaptive transpose Jacobian algorithms show
promising results even in higher mass uncertainties.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 17:04:53 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Shojaei",
"Hossein",
""
],
[
"Rahmanei",
"Hamid",
""
],
[
"Sadati",
"Seyed Hossein",
""
]
] |
2504.08727 | Boyang Deng | Boyang Deng, Songyou Peng, Kyle Genova, Gordon Wetzstein, Noah
Snavely, Leonidas Guibas, Thomas Funkhouser | Visual Chronicles: Using Multimodal LLMs to Analyze Massive Collections
of Images | Project page: https://boyangdeng.com/visual-chronicles , second and
third listed authors have equal contributions | null | null | null | cs.CV cs.AI cs.CY | http://creativecommons.org/licenses/by-sa/4.0/ | We present a system using Multimodal LLMs (MLLMs) to analyze a large database
with tens of millions of images captured at different times, with the aim of
discovering patterns in temporal changes. Specifically, we aim to capture
frequent co-occurring changes ("trends") across a city over a certain period.
Unlike previous visual analyses, our analysis answers open-ended queries (e.g.,
"what are the frequent types of changes in the city?") without any
predetermined target subjects or training labels. These properties cast prior
learning-based or unsupervised visual analysis tools unsuitable. We identify
MLLMs as a novel tool for their open-ended semantic understanding capabilities.
Yet, our datasets are four orders of magnitude too large for an MLLM to ingest
as context. So we introduce a bottom-up procedure that decomposes the massive
visual analysis problem into more tractable sub-problems. We carefully design
MLLM-based solutions to each sub-problem. During experiments and ablation
studies with our system, we find it significantly outperforms baselines and is
able to discover interesting trends from images captured in large cities (e.g.,
"addition of outdoor dining,", "overpass was painted blue," etc.). See more
results and interactive demos at https://boyangdeng.com/visual-chronicles.
| [
{
"version": "v1",
"created": "Fri, 11 Apr 2025 17:55:45 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 17:30:56 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Deng",
"Boyang",
""
],
[
"Peng",
"Songyou",
""
],
[
"Genova",
"Kyle",
""
],
[
"Wetzstein",
"Gordon",
""
],
[
"Snavely",
"Noah",
""
],
[
"Guibas",
"Leonidas",
""
],
[
"Funkhouser",
"Thomas",
""
]
] |
2504.08737 | Ben Rachmut | Ben Rachmut, Roie Zivan, William Yeoh | Latency-Aware 2-Opt Monotonic Local Search for Distributed Constraint
Optimization | null | In 30th International Conference on Principles and Practice of
Constraint Programming (CP 2024) (pp. 24-1) | 10.4230/LIPIcs.CP.2024.24 | null | cs.AI cs.DC | http://creativecommons.org/licenses/by/4.0/ | Researchers recently extended Distributed Constraint Optimization Problems
(DCOPs) to Communication-Aware DCOPs so that they are applicable in scenarios
in which messages can be arbitrarily delayed. Distributed asynchronous local
search and inference algorithms designed for CA-DCOPs are less vulnerable to
message latency than their counterparts for regular DCOPs. However, unlike
local search algorithms for (regular) DCOPs that converge to k-opt solutions
(with k > 1), that is, they converge to solutions that cannot be improved by a
group of k agents), local search CA-DCOP algorithms are limited to 1-opt
solutions only. In this paper, we introduce Latency-Aware Monotonic Distributed
Local Search-2 (LAMDLS-2), where agents form pairs and coordinate bilateral
assignment replacements. LAMDLS-2 is monotonic, converges to a 2-opt solution,
and is also robust to message latency, making it suitable for CA-DCOPs. Our
results indicate that LAMDLS-2 converges faster than MGM-2, a benchmark
algorithm, to a similar 2-opt solution, in various message latency scenarios.
| [
{
"version": "v1",
"created": "Fri, 21 Feb 2025 20:00:05 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Rachmut",
"Ben",
""
],
[
"Zivan",
"Roie",
""
],
[
"Yeoh",
"William",
""
]
] |
2504.08739 | Edward Sun | Edward Sun | Enhancing Product Search Interfaces with Sketch-Guided Diffusion and
Language Agents | Companion Proceedings of the ACM Web Conference 2025 | null | 10.1145/3701716.3717853 | null | cs.IR cs.HC | http://creativecommons.org/licenses/by/4.0/ | The rapid progress in diffusion models, transformers, and language agents has
unlocked new possibilities, yet their potential in user interfaces and
commercial applications remains underexplored. We present Sketch-Search Agent,
a novel framework that transforms the image search experience by integrating a
multimodal language agent with freehand sketches as control signals for
diffusion models. Using the T2I-Adapter, Sketch-Search Agent combines sketches
and text prompts to generate high-quality query images, encoded via a CLIP
image encoder for efficient matching against an image corpus. Unlike existing
methods, Sketch-Search Agent requires minimal setup, no additional training,
and excels in sketch-based image retrieval and natural language interactions.
The multimodal agent enhances user experience by dynamically retaining
preferences, ranking results, and refining queries for personalized
recommendations. This interactive design empowers users to create sketches and
receive tailored product suggestions, showcasing the potential of diffusion
models in user-centric image retrieval. Experiments confirm Sketch-Search
Agent's high accuracy in delivering relevant product search results.
| [
{
"version": "v1",
"created": "Fri, 21 Mar 2025 05:44:15 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Sun",
"Edward",
""
]
] |
2504.08740 | Yi Han | Kowei Shih, Yi Han, Li Tan | Recommendation System in Advertising and Streaming Media: Unsupervised
Data Enhancement Sequence Suggestions | null | null | null | null | cs.IR cs.LG | http://creativecommons.org/licenses/by/4.0/ | Sequential recommendation is an extensively explored approach to capturing
users' evolving preferences based on past interactions, aimed at predicting
their next likely choice. Despite significant advancements in this domain,
including methods based on RNNs and self-attention, challenges like limited
supervised signals and noisy data caused by unintentional clicks persist. To
address these challenges, some studies have incorporated unsupervised learning
by leveraging local item contexts within individual sequences. However, these
methods often overlook the intricate associations between items across multiple
sequences and are susceptible to noise in item co-occurrence patterns. In this
context, we introduce a novel framework, Global Unsupervised Data-Augmentation
(UDA4SR), which adopts a graph contrastive learning perspective to generate
more robust item embeddings for sequential recommendation. Our approach begins
by integrating Generative Adversarial Networks (GANs) for data augmentation,
which serves as the first step to enhance the diversity and richness of the
training data. Then, we build a Global Item Relationship Graph (GIG) based on
all user interaction sequences. Subsequently, we employ graph contrastive
learning on the refined graph to enhance item embeddings by capturing complex
global associations. To model users' dynamic and diverse interests more
effectively, we enhance the CapsNet module with a novel target-attention
mechanism. Extensive experiments show that UDA4SR significantly outperforms
state-of-the-art approaches.
| [
{
"version": "v1",
"created": "Sun, 23 Mar 2025 06:30:48 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Shih",
"Kowei",
""
],
[
"Han",
"Yi",
""
],
[
"Tan",
"Li",
""
]
] |
2504.08741 | Narges Mehran | Narges Mehran, Zahra Najafabadi Samani, Reza Farahani, Josef Hammer,
Dragi Kimovski | DEEP: Edge-based Dataflow Processing with Hybrid Docker Hub and Regional
Registries | 4 pages, three figures, IPDPSW 2025 | null | null | null | cs.DC | http://creativecommons.org/licenses/by-sa/4.0/ | Reducing energy consumption is essential to lessen greenhouse gas emissions,
conserve natural resources, and help mitigate the impacts of climate change. In
this direction, edge computing, a complementary technology to cloud computing,
extends computational capabilities closer to the data producers, enabling
energy-efficient and latency-sensitive service delivery for end users. To
properly manage data and microservice storage, expanding the Docker Hub
registry to the edge using an AWS S3-compatible MinIO-based object storage
service can reduce completion time and energy consumption. To address this, we
introduce Docker rEgistry-based Edge dataflow Processing (DEEP) to optimize the
energy consumption of microservice-based application deployments by focusing on
deployments from Docker Hub and MinIO-based regional registries and their
processing on edge devices. After applying nash equilibrium and benchmarking
the execution of two compute-intensive machine learning (ML) applications of
video and text processing, we compare energy consumption across three
deployment scenarios: exclusively from Docker Hub, exclusively from the
regional registry, and a hybrid method utilizing both. Experimental results
show that deploying 83% of text processing microservices from the regional
registry improves the energy consumption by 0.34% (18J) compared to
microservice deployments exclusively from Docker Hub.
| [
{
"version": "v1",
"created": "Sun, 23 Mar 2025 07:53:15 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mehran",
"Narges",
""
],
[
"Samani",
"Zahra Najafabadi",
""
],
[
"Farahani",
"Reza",
""
],
[
"Hammer",
"Josef",
""
],
[
"Kimovski",
"Dragi",
""
]
] |
2504.08742 | Nicholas Sukiennik | Nicholas Sukiennik, Haoyu Wang, Zailin Zeng, Chen Gao, Yong Li | Simulating Filter Bubble on Short-video Recommender System with Large
Language Model Agents | Submitted to IJCAI 2025 | null | null | null | cs.IR cs.AI | http://creativecommons.org/licenses/by/4.0/ | An increasing reliance on recommender systems has led to concerns about the
creation of filter bubbles on social media, especially on short video platforms
like TikTok. However, their formation is still not entirely understood due to
the complex dynamics between recommendation algorithms and user feedback. In
this paper, we aim to shed light on these dynamics using a large language
model-based simulation framework. Our work employs real-world short-video data
containing rich video content information and detailed user-agents to
realistically simulate the recommendation-feedback cycle. Through large-scale
simulations, we demonstrate that LLMs can replicate real-world user-recommender
interactions, uncovering key mechanisms driving filter bubble formation. We
identify critical factors, such as demographic features and category attraction
that exacerbate content homogenization. To mitigate this, we design and test
interventions including various cold-start and feedback weighting strategies,
showing measurable reductions in filter bubble effects. Our framework enables
rapid prototyping of recommendation strategies, offering actionable solutions
to enhance content diversity in real-world systems. Furthermore, we analyze how
LLM-inherent biases may propagate through recommendations, proposing safeguards
to promote equity for vulnerable groups, such as women and low-income
populations. By examining the interplay between recommendation and LLM agents,
this work advances a deeper understanding of algorithmic bias and provides
practical tools to promote inclusive digital spaces.
| [
{
"version": "v1",
"created": "Sun, 23 Mar 2025 10:35:58 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Sukiennik",
"Nicholas",
""
],
[
"Wang",
"Haoyu",
""
],
[
"Zeng",
"Zailin",
""
],
[
"Gao",
"Chen",
""
],
[
"Li",
"Yong",
""
]
] |
2504.08743 | Yang Yang | Yang Yang, Tong Zhang, Jian Wu, and Lijie Su | Dynamic Topic Analysis in Academic Journals using Convex Non-negative
Matrix Factorization Method | 11 pages, 7 figures, 6 tables | null | null | null | cs.IR cs.LG cs.SY eess.SY math.OC stat.AP | http://creativecommons.org/licenses/by/4.0/ | With the rapid advancement of large language models, academic topic
identification and topic evolution analysis are crucial for enhancing AI's
understanding capabilities. Dynamic topic analysis provides a powerful approach
to capturing and understanding the temporal evolution of topics in large-scale
datasets. This paper presents a two-stage dynamic topic analysis framework that
incorporates convex optimization to improve topic consistency, sparsity, and
interpretability. In Stage 1, a two-layer non-negative matrix factorization
(NMF) model is employed to extract annual topics and identify key terms. In
Stage 2, a convex optimization algorithm refines the dynamic topic structure
using the convex NMF (cNMF) model, further enhancing topic integration and
stability. Applying the proposed method to IEEE journal abstracts from 2004 to
2022 effectively identifies and quantifies emerging research topics, such as
COVID-19 and digital twins. By optimizing sparsity differences in the
clustering feature space between traditional and emerging research topics, the
framework provides deeper insights into topic evolution and ranking analysis.
Moreover, the NMF-cNMF model demonstrates superior stability in topic
consistency. At sparsity levels of 0.4, 0.6, and 0.9, the proposed approach
improves topic ranking stability by 24.51%, 56.60%, and 36.93%, respectively.
The source code (to be open after publication) is available at
https://github.com/meetyangyang/CDNMF.
| [
{
"version": "v1",
"created": "Sun, 23 Mar 2025 14:31:47 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yang",
"Yang",
""
],
[
"Zhang",
"Tong",
""
],
[
"Wu",
"Jian",
""
],
[
"Su",
"Lijie",
""
]
] |
2504.08744 | Esmail Gumaan Esm | Esmail Gumaan | ExpertRAG: Efficient RAG with Mixture of Experts -- Optimizing Context
Retrieval for Adaptive LLM Responses | 30 pages, 4 figures | null | null | null | cs.IR cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | ExpertRAG is a novel theoretical framework that integrates Mixture-of-Experts
(MoE) architectures with Retrieval Augmented Generation (RAG) to advance the
efficiency and accuracy of knowledge-intensive language modeling. We propose a
dynamic retrieval gating mechanism coupled with expert routing, enabling the
model to selectively consult an external knowledge store or rely on specialized
internal experts based on the query's needs. The paper lays out the theoretical
foundations of ExpertRAG, including a probabilistic formulation that treats
retrieval and expert selection as latent decisions, and mathematical
justifications for its efficiency in both computation and knowledge
utilization. We derive formulae to quantify the expected computational cost
savings from selective retrieval and the capacity gains from sparse expert
utilization. A comparative analysis positions ExpertRAG against standard RAG
(with always-on retrieval) and pure MoE models (e.g., Switch Transformer,
Mixtral) to highlight its unique balance between parametric knowledge and
non-parametric retrieval. We also outline an experimental validation strategy,
proposing benchmarks and evaluation protocols to test ExpertRAG's performance
on factual recall, generalization, and inference efficiency. The proposed
framework, although presented theoretically, is supported by insights from
prior work in RAG and MoE, and is poised to provide more factual, efficient,
and adaptive generation by leveraging the best of both paradigms. In summary,
ExpertRAG contributes a new perspective on scaling and augmenting language
models, backed by a thorough analysis and a roadmap for empirical validation.
| [
{
"version": "v1",
"created": "Sun, 23 Mar 2025 17:26:23 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Gumaan",
"Esmail",
""
]
] |
2504.08745 | Mert Yazan | Mert Yazan, Suzan Verberne, Frederik Situmeang | Improving RAG for Personalization with Author Features and Contrastive
Examples | null | null | null | null | cs.IR cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Personalization with retrieval-augmented generation (RAG) often fails to
capture fine-grained features of authors, making it hard to identify their
unique traits. To enrich the RAG context, we propose providing Large Language
Models (LLMs) with author-specific features, such as average sentiment polarity
and frequently used words, in addition to past samples from the author's
profile. We introduce a new feature called Contrastive Examples: documents from
other authors are retrieved to help LLM identify what makes an author's style
unique in comparison to others. Our experiments show that adding a couple of
sentences about the named entities, dependency patterns, and words a person
uses frequently significantly improves personalized text generation. Combining
features with contrastive examples boosts the performance further, achieving a
relative 15% improvement over baseline RAG while outperforming the benchmarks.
Our results show the value of fine-grained features for better personalization,
while opening a new research dimension for including contrastive examples as a
complement with RAG. We release our code publicly.
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 01:41:22 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yazan",
"Mert",
""
],
[
"Verberne",
"Suzan",
""
],
[
"Situmeang",
"Frederik",
""
]
] |
2504.08746 | Ngoc Luyen LE | Ngoc Luyen Le (Heudiasyc), Marie-H\'el\`ene Abel (Heudiasyc) | Enhancing Recommender Systems Using Textual Embeddings from Pre-trained
Language Models | null | The 8th International Conference on Information Technology &
Systems, Jan 2025, Mexico City, Mexico | null | null | cs.IR cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements in language models and pre-trained language models like
BERT and RoBERTa have revolutionized natural language processing, enabling a
deeper understanding of human-like language. In this paper, we explore
enhancing recommender systems using textual embeddings from pre-trained
language models to address the limitations of traditional recommender systems
that rely solely on explicit features from users, items, and user-item
interactions. By transforming structured data into natural language
representations, we generate high-dimensional embeddings that capture deeper
semantic relationships between users, items, and contexts. Our experiments
demonstrate that this approach significantly improves recommendation accuracy
and relevance, resulting in more personalized and context-aware
recommendations. The findings underscore the potential of PLMs to enhance the
effectiveness of recommender systems.
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 09:03:12 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Le",
"Ngoc Luyen",
"",
"Heudiasyc"
],
[
"Abel",
"Marie-Hélène",
"",
"Heudiasyc"
]
] |
2504.08747 | Jordan Chipka | Jordan Chipka, Chris Moyer, Clay Troyer, Tyler Fuelling, Jeremy
Hochstedler | GridMind: A Multi-Agent NLP Framework for Unified, Cross-Modal NFL Data
Insights | 16 pages, 2 figures, submitted to 2025 Sloan Sports Analytics
Conference | null | null | null | cs.AI cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid growth of big data and advancements in computational techniques
have significantly transformed sports analytics. However, the diverse range of
data sources -- including structured statistics, semi-structured formats like
sensor data, and unstructured media such as written articles, audio, and video
-- creates substantial challenges in extracting actionable insights. These
various formats, often referred to as multimodal data, require integration to
fully leverage their potential. Conventional systems, which typically
prioritize structured data, face limitations when processing and combining
these diverse content types, reducing their effectiveness in real-time sports
analysis.
To address these challenges, recent research highlights the importance of
multimodal data integration for capturing the complexity of real-world sports
environments. Building on this foundation, this paper introduces GridMind, a
multi-agent framework that unifies structured, semi-structured, and
unstructured data through Retrieval-Augmented Generation (RAG) and large
language models (LLMs) to facilitate natural language querying of NFL data.
This approach aligns with the evolving field of multimodal representation
learning, where unified models are increasingly essential for real-time,
cross-modal interactions.
GridMind's distributed architecture includes specialized agents that
autonomously manage each stage of a prompt -- from interpretation and data
retrieval to response synthesis. This modular design enables flexible, scalable
handling of multimodal data, allowing users to pose complex, context-rich
questions and receive comprehensive, intuitive responses via a conversational
interface.
| [
{
"version": "v1",
"created": "Mon, 24 Mar 2025 18:33:36 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Chipka",
"Jordan",
""
],
[
"Moyer",
"Chris",
""
],
[
"Troyer",
"Clay",
""
],
[
"Fuelling",
"Tyler",
""
],
[
"Hochstedler",
"Jeremy",
""
]
] |
2504.08748 | Lang Mei | Lang Mei, Siyu Mo, Zhihan Yang, Chong Chen | A Survey of Multimodal Retrieval-Augmented Generation | null | null | null | null | cs.IR cs.AI cs.CL cs.ET cs.LG | http://creativecommons.org/licenses/by/4.0/ | Multimodal Retrieval-Augmented Generation (MRAG) enhances large language
models (LLMs) by integrating multimodal data (text, images, videos) into
retrieval and generation processes, overcoming the limitations of text-only
Retrieval-Augmented Generation (RAG). While RAG improves response accuracy by
incorporating external textual knowledge, MRAG extends this framework to
include multimodal retrieval and generation, leveraging contextual information
from diverse data types. This approach reduces hallucinations and enhances
question-answering systems by grounding responses in factual, multimodal
knowledge. Recent studies show MRAG outperforms traditional RAG, especially in
scenarios requiring both visual and textual understanding. This survey reviews
MRAG's essential components, datasets, evaluation methods, and limitations,
providing insights into its construction and improvement. It also identifies
challenges and future research directions, highlighting MRAG's potential to
revolutionize multimodal information retrieval and generation. By offering a
comprehensive perspective, this work encourages further exploration into this
promising paradigm.
| [
{
"version": "v1",
"created": "Wed, 26 Mar 2025 02:43:09 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mei",
"Lang",
""
],
[
"Mo",
"Siyu",
""
],
[
"Yang",
"Zhihan",
""
],
[
"Chen",
"Chong",
""
]
] |
2504.08749 | Neal Graneau | Neal Graneau | Fundamental longitudinal electromagnetic (EM) force investigation using
DC current | 7 pages, 12 figures | null | null | null | physics.app-ph cond-mat.mtrl-sci physics.class-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The purpose of this work was to investigate historical claims of the
existence of a longitudinal ElectroMagnetic (EM) force component acting on
metallic atomic current elements in a direction parallel to the current flowing
through them. This lies outside conventional textbook physics predictions, yet
its existence has been indicated previously and if eventually confirmed will
have a significant effect on physics theory and many technological
applications, especially those involving high current density (> 10^9 A/m2).
The experiment described here is based on the measurement of force on a copper
armature submerged in a trough containing liquid metal through which constant
DC current is passing. The coaxial symmetry of the experiment (CRE Coaxial
Recoil Experiment) was able to limit all net force on the centred armature to
the direction of interest, parallel to the current within it. All the
experimental data related the measured forces on the armature to the current
flowing through the circuit. Variations were made to the length and location of
the armature along the central axis as well as modifications to the liquid
metal configuration. Raw and processed results are presented, and an
experimental technique described that revealed strong evidence for the
existence of and discrimination between axial mechanical contact forces and
longitudinal EM force. The original EM force law, proposed in 1822 by Ampere,
includes a longitudinal component and has been found to be qualitatively
consistent with all experiments to date, including these reported findings, and
is considered a candidate explanatory theory.
| [
{
"version": "v1",
"created": "Wed, 26 Mar 2025 13:35:27 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Graneau",
"Neal",
""
]
] |
2504.08750 | Biswanath Dutta Dr. | Biswanath Dutta and Debanjali Bain | A Quantitative Approach to Evaluating Open-Source EHR Systems for Indian
Healthcare | 21 pages, 3 figures, 10 tables | null | null | null | cs.IR cs.DL | http://creativecommons.org/licenses/by/4.0/ | The increasing use of Electronic Health Records (EHR) has emphasized the need
for standardization and interoperability in healthcare data management. The
Ministry of Health and Family Welfare, Government of India, has introduced the
Electronic Health Record Minimum Data Set (EHRMDS) to facilitate uniformity in
clinical documentation. However, the compatibility of Open-Source Electronic
Health Record Systems (OS-EHRS) with EHRMDS remains largely unexplored. This
study conducts a systematic assessment of the alignment between EHRMDS and
commonly utilized OS-EHRS to determine the most appropriate system for
healthcare environments in India. A quantitative closeness analysis was
performed by comparing the metadata elements of EHRMDS with those of 10
selected OS-EHRS. Using crosswalk methodologies based on syntactic and semantic
similarity, the study measured the extent of metadata alignment. Results
indicate that OpenEMR exhibits the highest compatibility with EHRMDS, covering
73.81% of its metadata elements, while OpenClinic shows the least alignment at
33.33%. Additionally, the analysis identified 47 metadata elements present in
OS-EHRS but absent in EHRMDS, suggesting the need for an extended metadata
schema. By bridging gaps in clinical metadata, this study contributes to
enhancing the interoperability of EHR systems in India. The findings provide
valuable insights for healthcare policymakers and organizations seeking to
adopt OS-EHRS aligned with national standards. Keywords. EHR metadata,
electronic health record systems, EHRMDS, meta data, structured vocabularies,
metadata crosswalk, methodologies and tools, SNOMED-CT, UMLS terms.
| [
{
"version": "v1",
"created": "Thu, 27 Mar 2025 14:55:03 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Dutta",
"Biswanath",
""
],
[
"Bain",
"Debanjali",
""
]
] |
2504.08751 | Qingyi Lu | Haowei Yang, Lei Fu, Qingyi Lu, Yue Fan, Tianle Zhang, Ruohan Wang | Research on the Design of a Short Video Recommendation System Based on
Multimodal Information and Differential Privacy | null | null | null | null | cs.IR cs.AI cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the rapid development of short video platforms, recommendation systems
have become key technologies for improving user experience and enhancing
platform engagement. However, while short video recommendation systems leverage
multimodal information (such as images, text, and audio) to improve
recommendation effectiveness, they also face the severe challenge of user
privacy leakage. This paper proposes a short video recommendation system based
on multimodal information and differential privacy protection. First, deep
learning models are used for feature extraction and fusion of multimodal data,
effectively improving recommendation accuracy. Then, a differential privacy
protection mechanism suitable for recommendation scenarios is designed to
ensure user data privacy while maintaining system performance. Experimental
results show that the proposed method outperforms existing mainstream
approaches in terms of recommendation accuracy, multimodal fusion
effectiveness, and privacy protection performance, providing important insights
for the design of recommendation systems for short video platforms.
| [
{
"version": "v1",
"created": "Thu, 27 Mar 2025 22:56:41 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yang",
"Haowei",
""
],
[
"Fu",
"Lei",
""
],
[
"Lu",
"Qingyi",
""
],
[
"Fan",
"Yue",
""
],
[
"Zhang",
"Tianle",
""
],
[
"Wang",
"Ruohan",
""
]
] |
2504.08752 | David Brett | David Brett and Anniek Myatt | Patience is all you need! An agentic system for performing scientific
literature review | 10 pages, 5 figures | null | null | null | cs.IR cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Large language models (LLMs) have grown in their usage to provide support for
question answering across numerous disciplines. The models on their own have
already shown promise for answering basic questions, however fail quickly where
expert domain knowledge is required or the question is nuanced. Scientific
research often involves searching for relevant literature, distilling pertinent
information from that literature and analysing how the findings support or
contradict one another. The information is often encapsulated in the full text
body of research articles, rather than just in the abstracts. Statements within
these articles frequently require the wider article context to be fully
understood. We have built an LLM-based system that performs such search and
distillation of information encapsulated in scientific literature, and we
evaluate our keyword based search and information distillation system against a
set of biology related questions from previously released literature
benchmarks. We demonstrate sparse retrieval methods exhibit results close to
state of the art without the need for dense retrieval, with its associated
infrastructure and complexity overhead. We also show how to increase the
coverage of relevant documents for literature review generation.
| [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 08:08:46 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Brett",
"David",
""
],
[
"Myatt",
"Anniek",
""
]
] |
2504.08753 | Satyanarayana Murthy | Jyothi, T. Satyanarayana Murthy | Domain Specific Question to SQL Conversion with Embedded Data Balancing
Technique | null | null | null | null | cs.IR cs.CL cs.DB | http://creativecommons.org/publicdomain/zero/1.0/ | The rise of deep learning in natural language processing has fostered the
creation of text to structured query language models composed of an encoder and
a decoder. Researchers have experimented with various intermediate processing
like schema linking, table type aware, value extract. To generate accurate SQL
results for the user question. However error analysis performed on the failed
cases on these systems shows, 29 percentage of the errors would be because the
system was unable to understand the values expressed by the user in their
question. This challenge affects the generation of accurate SQL queries,
especially when dealing with domain-specific terms and specific value
conditions, where traditional methods struggle to maintain consistency and
precision. To overcome these obstacles, proposed two intermediations like
implementing data balancing technique and over sampling domain-specific queries
which would refine the model architecture to enhance value recognition and fine
tuning the model for domain-specific questions. This proposed solution achieved
10.98 percentage improvement in accuracy of the model performance compared to
the state of the art model tested on WikiSQL dataset. to convert the user
question accurately to SQL queries. Applying oversampling technique on the
domain-specific questions shown a significant improvement as compared with
traditional approaches.
| [
{
"version": "v1",
"created": "Fri, 28 Mar 2025 08:58:14 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Jyothi",
"",
""
],
[
"Murthy",
"T. Satyanarayana",
""
]
] |
2504.08755 | Dirk Spennemann | Dirk HR Spennemann | Delving into: the quantification of Ai-generated content on the internet
(synthetic data) | 9 pp | null | null | null | cs.IR cs.AI cs.HC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | While it is increasingly evident that the internet is becoming saturated with
content created by generated Ai large language models, accurately measuring the
scale of this phenomenon has proven challenging. By analyzing the frequency of
specific keywords commonly used by ChatGPT, this paper demonstrates that such
linguistic markers can effectively be used to esti-mate the presence of
generative AI content online. The findings suggest that at least 30% of text on
active web pages originates from AI-generated sources, with the actual
proportion likely ap-proaching 40%. Given the implications of autophagous
loops, this is a sobering realization.
| [
{
"version": "v1",
"created": "Sat, 29 Mar 2025 03:06:53 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Spennemann",
"Dirk HR",
""
]
] |
2504.08756 | Kyohoon Jin | Jeongsoo Lee, Daeyong Kwon, Kyohoon Jin, Junnyeong Jeong, Minwoo Sim,
Minwoo Kim | MHTS: Multi-Hop Tree Structure Framework for Generating
Difficulty-Controllable QA Datasets for RAG Evaluation | null | null | null | null | cs.IR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Existing RAG benchmarks often overlook query difficulty, leading to inflated
performance on simpler questions and unreliable evaluations. A robust benchmark
dataset must satisfy three key criteria: quality, diversity, and difficulty,
which capturing the complexity of reasoning based on hops and the distribution
of supporting evidence. In this paper, we propose MHTS (Multi-Hop Tree
Structure), a novel dataset synthesis framework that systematically controls
multi-hop reasoning complexity by leveraging a multi-hop tree structure to
generate logically connected, multi-chunk queries. Our fine-grained difficulty
estimation formula exhibits a strong correlation with the overall performance
metrics of a RAG system, validating its effectiveness in assessing both
retrieval and answer generation capabilities. By ensuring high-quality,
diverse, and difficulty-controlled queries, our approach enhances RAG
evaluation and benchmarking capabilities.
| [
{
"version": "v1",
"created": "Sat, 29 Mar 2025 06:26:01 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lee",
"Jeongsoo",
""
],
[
"Kwon",
"Daeyong",
""
],
[
"Jin",
"Kyohoon",
""
],
[
"Jeong",
"Junnyeong",
""
],
[
"Sim",
"Minwoo",
""
],
[
"Kim",
"Minwoo",
""
]
] |
2504.08757 | Tiago Machado | Tiago Machado, Sara E. Berger, Cassia Sanctos, Vagner Figueiredo de
Santana, Lemara Williams, Zhaoqing Wu | A Framework for Lightweight Responsible Prompting Recommendation | 13 pages, 3 figures, 3 tables, 1 algorithm | null | null | null | cs.IR cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Computer Science and Design practitioners have been researching and proposing
alternatives for a dearth of recommendations, standards, or best practices in
user interfaces for decades. Now, with the advent of generative Artificial
Intelligence (GenAI), we have yet again an emerging, powerful technology that
lacks sufficient guidance in terms of possible interactions, inputs, and
outcomes. In this context, this work proposes a lightweight framework for
responsible prompting recommendation to be added before the prompt is sent to
GenAI. The framework is comprised of (1) a human-curated dataset for
recommendations, (2) a red team dataset for assessing recommendations, (3) a
sentence transformer for semantics mapping, (4) a similarity metric to map
input prompt to recommendations, (5) a set of similarity thresholds, (6)
quantized sentence embeddings, (7) a recommendation engine, and (8) an
evaluation step to use the red team dataset. With the proposed framework and
open-source system, the contributions presented can be applied in multiple
contexts where end-users can benefit from guidance for interacting with GenAI
in a more responsible way, recommending positive values to be added and harmful
sentences to be removed.
| [
{
"version": "v1",
"created": "Sat, 29 Mar 2025 13:56:49 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Machado",
"Tiago",
""
],
[
"Berger",
"Sara E.",
""
],
[
"Sanctos",
"Cassia",
""
],
[
"de Santana",
"Vagner Figueiredo",
""
],
[
"Williams",
"Lemara",
""
],
[
"Wu",
"Zhaoqing",
""
]
] |
2504.08758 | Yifan Feng | Yifan Feng, Hao Hu, Xingliang Hou, Shiquan Liu, Shihui Ying, Shaoyi
Du, Han Hu, Yue Gao | Hyper-RAG: Combating LLM Hallucinations using Hypergraph-Driven
Retrieval-Augmented Generation | null | null | null | null | cs.IR cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) have transformed various sectors, including
education, finance, and medicine, by enhancing content generation and
decision-making processes. However, their integration into the medical field is
cautious due to hallucinations, instances where generated content deviates from
factual accuracy, potentially leading to adverse outcomes. To address this, we
introduce Hyper-RAG, a hypergraph-driven Retrieval-Augmented Generation method
that comprehensively captures both pairwise and beyond-pairwise correlations in
domain-specific knowledge, thereby mitigating hallucinations. Experiments on
the NeurologyCrop dataset with six prominent LLMs demonstrated that Hyper-RAG
improves accuracy by an average of 12.3% over direct LLM use and outperforms
Graph RAG and Light RAG by 6.3% and 6.0%, respectively. Additionally, Hyper-RAG
maintained stable performance with increasing query complexity, unlike existing
methods which declined. Further validation across nine diverse datasets showed
a 35.5% performance improvement over Light RAG using a selection-based
assessment. The lightweight variant, Hyper-RAG-Lite, achieved twice the
retrieval speed and a 3.3% performance boost compared with Light RAG. These
results confirm Hyper-RAG's effectiveness in enhancing LLM reliability and
reducing hallucinations, making it a robust solution for high-stakes
applications like medical diagnostics.
| [
{
"version": "v1",
"created": "Sun, 30 Mar 2025 12:39:14 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Feng",
"Yifan",
""
],
[
"Hu",
"Hao",
""
],
[
"Hou",
"Xingliang",
""
],
[
"Liu",
"Shiquan",
""
],
[
"Ying",
"Shihui",
""
],
[
"Du",
"Shaoyi",
""
],
[
"Hu",
"Han",
""
],
[
"Gao",
"Yue",
""
]
] |
2504.08759 | Cristinel Diaconu | Yasmine Amhis, Jeremy Andrea, Etienne Aug\'e, Sara Bolognesi, Maarten
Boonekamp, Samuel Calvet, Emilien Chapon, Didier Contardo, Fabrice Couderc,
Sabine Cr\'ep\'e-Renaudin, Louis D'Eramo, Cristinel Diaconu, Giulio Dujany,
Federico Ferri, Marie-H\'el\`ene Genest, St\'ephane Lavignac, Jessica
Lev\^eque, Cyrille Marquet, Anselmo Meregaglia, Stephane Monteil, Carlos
Mu\~noz Camacho, Louis Portales, Philippe Schwemling, Christopher Smith, Ana
M. Teixeira, Michael Winn | French HEP community input to the European Strategy for Particle Physics | Contribution to the European Strategy for Particle Physics 2026
(cover page + 10 pages ; no references) | null | null | null | hep-ex hep-lat hep-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | In view of the European Strategy for Particle Physics process, the French HEP
community has organized a national process of collecting written contributions
and has pursued a series of workshops culminating with a national symposium
held in Paris on January 20-21, 2025 that involved over 280 scientists
https://indico.in2p3.fr/event/34662/. The present document summarises the main
conclusions of this bottom-up approach centred on the physics and technology
motivations.
| [
{
"version": "v1",
"created": "Sun, 30 Mar 2025 22:54:57 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Amhis",
"Yasmine",
""
],
[
"Andrea",
"Jeremy",
""
],
[
"Augé",
"Etienne",
""
],
[
"Bolognesi",
"Sara",
""
],
[
"Boonekamp",
"Maarten",
""
],
[
"Calvet",
"Samuel",
""
],
[
"Chapon",
"Emilien",
""
],
[
"Contardo",
"Didier",
""
],
[
"Couderc",
"Fabrice",
""
],
[
"Crépé-Renaudin",
"Sabine",
""
],
[
"D'Eramo",
"Louis",
""
],
[
"Diaconu",
"Cristinel",
""
],
[
"Dujany",
"Giulio",
""
],
[
"Ferri",
"Federico",
""
],
[
"Genest",
"Marie-Hélène",
""
],
[
"Lavignac",
"Stéphane",
""
],
[
"Levêque",
"Jessica",
""
],
[
"Marquet",
"Cyrille",
""
],
[
"Meregaglia",
"Anselmo",
""
],
[
"Monteil",
"Stephane",
""
],
[
"Camacho",
"Carlos Muñoz",
""
],
[
"Portales",
"Louis",
""
],
[
"Schwemling",
"Philippe",
""
],
[
"Smith",
"Christopher",
""
],
[
"Teixeira",
"Ana M.",
""
],
[
"Winn",
"Michael",
""
]
] |
2504.08760 | Yi-Cai Zhang | Guang-Xin Pang, Zhi Li, Shan-Zhong Li, Yan-Yang Zhang, Jun-Feng Liu,
and Yi-Cai Zhang | Exact mobility line and mobility ring in the complex energy plane of a
flat band lattice with a non-Hermitian quasiperiodic potential | 13 pages, 9 figures | null | null | null | cond-mat.dis-nn cond-mat.quant-gas | http://creativecommons.org/publicdomain/zero/1.0/ | In this study, we investigate the problem of Anderson localization in a
one-dimensional flat band lattice with a non-Hermitian quasiperiodic on-site
potential. First of all, we discuss the influences of non-Hermitian potentials
on the existence of critical states. Our findings show that, unlike in
Hermitian cases, the non-Hermiticity of the potential leads to the
disappearance of critical states and critical regions. Furthermore, we are able
to accurately determine the Lyapunov exponents and the mobility edges. Our
results reveal that the mobility edges form mobility lines and mobility rings
in the complex energy plane. Within the mobility rings, the eigenstates are
extended, while the localized states are located outside the mobility rings.
For mobility line cases, only when the eigenenergies lie on the mobility lines,
their corresponding eigenstates are extended states.Finally, as the energy
approaches the mobility edges, we observe that, differently from Hermitian
cases, here the critical index of the localization length is not a constant,
but rather varies depending on the positions of the mobility edges.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 01:55:12 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Pang",
"Guang-Xin",
""
],
[
"Li",
"Zhi",
""
],
[
"Li",
"Shan-Zhong",
""
],
[
"Zhang",
"Yan-Yang",
""
],
[
"Liu",
"Jun-Feng",
""
],
[
"Zhang",
"Yi-Cai",
""
]
] |
2504.08761 | Sen Mei | Yuxuan Chen, Dewen Guo, Sen Mei, Xinze Li, Hao Chen, Yishan Li, Yixuan
Wang, Chaoyue Tang, Ruobing Wang, Dingjun Wu, Yukun Yan, Zhenghao Liu, Shi
Yu, Zhiyuan Liu, Maosong Sun | UltraRAG: A Modular and Automated Toolkit for Adaptive
Retrieval-Augmented Generation | null | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Retrieval-Augmented Generation (RAG) significantly enhances the performance
of large language models (LLMs) in downstream tasks by integrating external
knowledge. To facilitate researchers in deploying RAG systems, various RAG
toolkits have been introduced. However, many existing RAG toolkits lack support
for knowledge adaptation tailored to specific application scenarios. To address
this limitation, we propose UltraRAG, a RAG toolkit that automates knowledge
adaptation throughout the entire workflow, from data construction and training
to evaluation, while ensuring ease of use. UltraRAG features a user-friendly
WebUI that streamlines the RAG process, allowing users to build and optimize
systems without coding expertise. It supports multimodal input and provides
comprehensive tools for managing the knowledge base. With its highly modular
architecture, UltraRAG delivers an end-to-end development solution, enabling
seamless knowledge adaptation across diverse user scenarios. The code,
demonstration videos, and installable package for UltraRAG are publicly
available at https://github.com/OpenBMB/UltraRAG.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 03:49:49 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Chen",
"Yuxuan",
""
],
[
"Guo",
"Dewen",
""
],
[
"Mei",
"Sen",
""
],
[
"Li",
"Xinze",
""
],
[
"Chen",
"Hao",
""
],
[
"Li",
"Yishan",
""
],
[
"Wang",
"Yixuan",
""
],
[
"Tang",
"Chaoyue",
""
],
[
"Wang",
"Ruobing",
""
],
[
"Wu",
"Dingjun",
""
],
[
"Yan",
"Yukun",
""
],
[
"Liu",
"Zhenghao",
""
],
[
"Yu",
"Shi",
""
],
[
"Liu",
"Zhiyuan",
""
],
[
"Sun",
"Maosong",
""
]
] |
2504.08762 | Zhiyuan Wen | Zhiyuan Wen, Jiannong Cao, Zian Wang, Beichen Guo, Ruosong Yang,
Shuaiqi Liu | InteractiveSurvey: An LLM-based Personalized and Interactive Survey
Paper Generation System | null | null | null | null | cs.IR cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The exponential growth of academic literature creates urgent demands for
comprehensive survey papers, yet manual writing remains time-consuming and
labor-intensive. Recent advances in large language models (LLMs) and
retrieval-augmented generation (RAG) facilitate studies in synthesizing survey
papers from multiple references, but most existing works restrict users to
title-only inputs and fixed outputs, neglecting the personalized process of
survey paper writing. In this paper, we introduce InteractiveSurvey - an
LLM-based personalized and interactive survey paper generation system.
InteractiveSurvey can generate structured, multi-modal survey papers with
reference categorizations from multiple reference papers through both online
retrieval and user uploads. More importantly, users can customize and refine
intermediate components continuously during generation, including reference
categorization, outline, and survey content through an intuitive interface.
Evaluations of content quality, time efficiency, and user studies show that
InteractiveSurvey is an easy-to-use survey generation system that outperforms
most LLMs and existing methods in output content quality while remaining highly
time-efficient.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 04:23:22 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wen",
"Zhiyuan",
""
],
[
"Cao",
"Jiannong",
""
],
[
"Wang",
"Zian",
""
],
[
"Guo",
"Beichen",
""
],
[
"Yang",
"Ruosong",
""
],
[
"Liu",
"Shuaiqi",
""
]
] |
2504.08763 | Mario M. Kubek | Shiraj Pokharel, Georg P. Ro{\ss}rucker, Mario M. Kubek | WebMap -- Large Language Model-assisted Semantic Link Induction in the
Web | 11 pages, 3 figures, accepted at the 2024 24th International
Conference on Innovations for Community Services (I4CS), June 12 - 14,
Maastricht, The Netherlands, 2024 | null | 10.1007/978-3-031-60433-1_8 | null | cs.IR cs.AI cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Carrying out research tasks is only inadequately supported, if not hindered,
by current web search engines. This paper therefore proposes functional
extensions of WebMap, a semantically induced overlay linking structure on the
web to inherently facilitate research activities. These add-ons support the
dynamic determination and regrouping of document clusters, the creation of a
semantic signpost in the web, and the interactive tracing of topics back to
their origins.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 04:40:45 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Pokharel",
"Shiraj",
""
],
[
"Roßrucker",
"Georg P.",
""
],
[
"Kubek",
"Mario M.",
""
]
] |
2504.08764 | Chris Brogly | Chris Brogly, Saif Rjaibi, Charlotte Liang, Erica Lam, Edward Wang,
Adam Levitan, Sarah Paleczny, Michael Cusimano | Evaluation of the phi-3-mini SLM for identification of texts related to
medicine, health, and sports injuries | null | null | null | null | cs.IR cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Small Language Models (SLMs) have potential to be used for automatically
labelling and identifying aspects of text data for medicine/health-related
purposes from documents and the web. As their resource requirements are
significantly lower than Large Language Models (LLMs), these can be deployed
potentially on more types of devices. SLMs often are benchmarked on
health/medicine-related tasks, such as MedQA, although performance on these can
vary especially depending on the size of the model in terms of number of
parameters. Furthermore, these test results may not necessarily reflect
real-world performance regarding the automatic labelling or identification of
texts in documents and the web. As a result, we compared topic-relatedness
scores from Microsofts phi-3-mini-4k-instruct SLM to the topic-relatedness
scores from 7 human evaluators on 1144 samples of medical/health-related texts
and 1117 samples of sports injury-related texts. These texts were from a larger
dataset of about 9 million news headlines, each of which were processed and
assigned scores by phi-3-mini-4k-instruct. Our sample was selected (filtered)
based on 1 (low filtering) or more (high filtering) Boolean conditions on the
phi-3 SLM scores. We found low-moderate significant correlations between the
scores from the SLM and human evaluators for sports injury texts with low
filtering (\r{ho} = 0.3413, p < 0.001) and medicine/health texts with high
filtering (\r{ho} = 0.3854, p < 0.001), and low significant correlation for
medicine/health texts with low filtering (\r{ho} = 0.2255, p < 0.001). There
was negligible, insignificant correlation for sports injury-related texts with
high filtering (\r{ho} = 0.0318, p = 0.4466).
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 07:25:59 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Brogly",
"Chris",
""
],
[
"Rjaibi",
"Saif",
""
],
[
"Liang",
"Charlotte",
""
],
[
"Lam",
"Erica",
""
],
[
"Wang",
"Edward",
""
],
[
"Levitan",
"Adam",
""
],
[
"Paleczny",
"Sarah",
""
],
[
"Cusimano",
"Michael",
""
]
] |
2504.08765 | Sepideh Khodaparast | Bashra Mahamed, Francis James Dent, Robert Simpson, Nicola Weston,
Fanny Nascimento Costa, and Sepideh Khodaparast | Dew harvesting grass: role of epicuticular wax in regulating
condensation dynamics | null | null | null | null | cond-mat.soft | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Identification and characterization of natural dew collecting models is
instrumental for the inspiration, design and development of engineered dew
harvesting systems. Short low growing grass is one of the most ubiquitous and
proficient examples of natural dew harvesting, owing to its large surface area,
small thermal capacity, structured rough surface and proximity to ground level.
Here, we provide a closer look at the formation, growth, and dynamics of
microscale dew droplets on the surface of wheatgrass leaves, investigating the
role of epicuticular wax. The wheatgrass leaf exhibits biphilic properties
emerging from the hydrophilic lamina covered by hydrophobic wax
microsculptures. As a result, the regulation of the dew formation and
condensation dynamics is largely governed by the arrangement and density of
epicuticular wax micromorphologies. At moderate subcooling levels (4-10
$^{\circ}$C below the dew point), we observe drop-wise condensation on the
superhydrophobic adaxial side, while significant flooding and film condensation
usually appear on abaxial surfaces with lower wax coverage. On the adaxial side
of the leaves, the fairly uniform coverage of the hydrophobic epicuticular wax
crystals on the hydrophilic background promotes drop-wise condensation
nucleation while facilitating droplet mobility. Frequent coalescence of
multiple droplets of 5 - 20 ${\mu}$m diameter results in self-propelled
departure events, creating free potential sites for new nucleation. The
findings of this study advance our understanding of dew formation on natural
surfaces while providing inspiration and guidance for the development of
sustainable functional microstructured coatings for various drop-wise
condensation applications.
| [
{
"version": "v1",
"created": "Mon, 31 Mar 2025 14:44:11 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mahamed",
"Bashra",
""
],
[
"Dent",
"Francis James",
""
],
[
"Simpson",
"Robert",
""
],
[
"Weston",
"Nicola",
""
],
[
"Costa",
"Fanny Nascimento",
""
],
[
"Khodaparast",
"Sepideh",
""
]
] |
2504.08766 | Hongyang Cheng | Marc Fransen, Andreas F\"urst, Deepak Tunuguntla, Daniel N. Wilke,
Benedikt Alkin, Daniel Barreto, Johannes Brandstetter, Miguel Angel Cabrera,
Xinyan Fan, Mengwu Guo, Bram Kieskamp, Krishna Kumar, John Morrissey,
Jonathan Nuttall, Jin Ooi, Luisa Orozco, Stefanos-Aldo Papanicolopulos,
Tongming Qu, Dingena Schott, Takayuki Shuku, WaiChing Sun, Thomas Weinhart,
Dongwei Ye, Hongyang Cheng | Towards scientific machine learning for granular material simulations --
challenges and opportunities | 35 pages, 17 figures | null | null | null | cond-mat.soft cs.LG physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | Micro-scale mechanisms, such as inter-particle and particle-fluid
interactions, govern the behaviour of granular systems. While particle-scale
simulations provide detailed insights into these interactions, their
computational cost is often prohibitive. Attended by researchers from both the
granular materials (GM) and machine learning (ML) communities, a recent Lorentz
Center Workshop on "Machine Learning for Discrete Granular Media" brought the
ML community up to date with GM challenges.
This position paper emerged from the workshop discussions. We define granular
materials and identify seven key challenges that characterise their distinctive
behaviour across various scales and regimes, ranging from gas-like to
fluid-like and solid-like. Addressing these challenges is essential for
developing robust and efficient digital twins for granular systems in various
industrial applications. To showcase the potential of ML to the GM community,
we present classical and emerging machine/deep learning techniques that have
been, or could be, applied to granular materials. We reviewed sequence-based
learning models for path-dependent constitutive behaviour, followed by
encoder-decoder type models for representing high-dimensional data. We then
explore graph neural networks and recent advances in neural operator learning.
Lastly, we discuss model-order reduction and probabilistic learning techniques
for high-dimensional parameterised systems, which are crucial for quantifying
uncertainties arising from physics-based and data-driven models.
We present a workflow aimed at unifying data structures and modelling
pipelines and guiding readers through the selection, training, and deployment
of ML surrogates for granular material simulations. Finally, we illustrate the
workflow's practical use with two representative examples, focusing on granular
materials in solid-like and fluid-like regimes.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 09:03:59 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Fransen",
"Marc",
""
],
[
"Fürst",
"Andreas",
""
],
[
"Tunuguntla",
"Deepak",
""
],
[
"Wilke",
"Daniel N.",
""
],
[
"Alkin",
"Benedikt",
""
],
[
"Barreto",
"Daniel",
""
],
[
"Brandstetter",
"Johannes",
""
],
[
"Cabrera",
"Miguel Angel",
""
],
[
"Fan",
"Xinyan",
""
],
[
"Guo",
"Mengwu",
""
],
[
"Kieskamp",
"Bram",
""
],
[
"Kumar",
"Krishna",
""
],
[
"Morrissey",
"John",
""
],
[
"Nuttall",
"Jonathan",
""
],
[
"Ooi",
"Jin",
""
],
[
"Orozco",
"Luisa",
""
],
[
"Papanicolopulos",
"Stefanos-Aldo",
""
],
[
"Qu",
"Tongming",
""
],
[
"Schott",
"Dingena",
""
],
[
"Shuku",
"Takayuki",
""
],
[
"Sun",
"WaiChing",
""
],
[
"Weinhart",
"Thomas",
""
],
[
"Ye",
"Dongwei",
""
],
[
"Cheng",
"Hongyang",
""
]
] |
2504.08767 | Bryar Hassan Dr. | Bryar A. Hassan, Alla A. Hassan, Joan Lu, Aram M. Ahmed, Tarik A.
Rashid | A Proposed Hybrid Recommender System for Tourism Industry in Iraq Using
Evolutionary Apriori and K-means Algorithms | null | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | The rapid proliferation of tourism data across sectors, including
accommodations, cultural sites, and events, has made it increasingly
challenging for travelers to identify relevant and personalized
recommendations. While traditional recommender systems such as collaborative,
content-based, and context-aware systems offer partial solutions, they often
struggle with issues like data sparsity and overspecialization. This study
proposes a novel hybrid recommender system that combines evolutionary Apriori
and K-means clustering algorithms to improve recommendation accuracy and
efficiency in the tourism domain. Designed specifically to address the diverse
and dynamic tourism landscape in Iraq, the system provides personalized
recommendations and clusters of tourist destinations tailored to user
preferences and contextual information. To evaluate the systems performance,
experiments were conducted on an augmented dataset representative of Iraqs
tourism activity, comparing the proposed system with existing methods. Results
indicate that the proposed hybrid system significantly reduces execution time
by 27-56% and space consumption by 24-31%, while achieving consistently lower
Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) values, thereby
enhancing prediction accuracy. This approach offers a scalable, context-aware
framework that is well-suited for application in regions where tourism data is
limited, such as Iraq, ultimately advancing tourism recommender systems by
addressing their limitations in complex and data-scarce environments.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 09:28:13 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hassan",
"Bryar A.",
""
],
[
"Hassan",
"Alla A.",
""
],
[
"Lu",
"Joan",
""
],
[
"Ahmed",
"Aram M.",
""
],
[
"Rashid",
"Tarik A.",
""
]
] |
2504.08768 | Xiaofan Zhou | Xiaofan Zhou, Liangjie Huang, Pinyang Cheng, Wenpen Yin, Rui Zhang,
Wenrui Hao, Lu Cheng | Accelerating Causal Network Discovery of Alzheimer Disease Biomarkers
via Scientific Literature-based Retrieval Augmented Generation | 9 pages, under review | null | null | null | cs.IR q-bio.QM | http://creativecommons.org/licenses/by/4.0/ | The causal relationships between biomarkers are essential for disease
diagnosis and medical treatment planning. One notable application is
Alzheimer's disease (AD) diagnosis, where certain biomarkers may influence the
presence of others, enabling early detection, precise disease staging, targeted
treatments, and improved monitoring of disease progression. However,
understanding these causal relationships is complex and requires extensive
research. Constructing a comprehensive causal network of biomarkers demands
significant effort from human experts, who must analyze a vast number of
research papers, and have bias in understanding diseases' biomarkers and their
relation. This raises an important question: Can advanced large language models
(LLMs), such as those utilizing retrieval-augmented generation (RAG), assist in
building causal networks of biomarkers for further medical analysis? To explore
this, we collected 200 AD-related research papers published over the past 25
years and then integrated scientific literature with RAG to extract AD
biomarkers and generate causal relations among them. Given the high-risk nature
of the medical diagnosis, we applied uncertainty estimation to assess the
reliability of the generated causal edges and examined the faithfulness and
scientificness of LLM reasoning using both automatic and human evaluation. We
find that RAG enhances the ability of LLMs to generate more accurate causal
networks from scientific papers. However, the overall performance of LLMs in
identifying causal relations of AD biomarkers is still limited. We hope this
study will inspire further foundational research on AI-driven analysis of AD
biomarkers causal network discovery.
| [
{
"version": "v1",
"created": "Tue, 1 Apr 2025 22:15:17 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhou",
"Xiaofan",
""
],
[
"Huang",
"Liangjie",
""
],
[
"Cheng",
"Pinyang",
""
],
[
"Yin",
"Wenpen",
""
],
[
"Zhang",
"Rui",
""
],
[
"Hao",
"Wenrui",
""
],
[
"Cheng",
"Lu",
""
]
] |
2504.08769 | Dario Izzo | Dario Izzo, Sebastien Origer, Giacomo Acciarini and Francesco Biscani | High-order expansion of Neural Ordinary Differential Equations flows | null | null | null | null | math.OC cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Artificial neural networks, widely recognised for their role in machine
learning, are now transforming the study of ordinary differential equations
(ODEs), bridging data-driven modelling with classical dynamical systems and
enabling the development of infinitely deep neural models. However, the
practical applicability of these models remains constrained by the opacity of
their learned dynamics, which operate as black-box systems with limited
explainability, thereby hindering trust in their deployment. Existing
approaches for the analysis of these dynamical systems are predominantly
restricted to first-order gradient information due to computational
constraints, thereby limiting the depth of achievable insight. Here, we
introduce Event Transition Tensors, a framework based on high-order
differentials that provides a rigorous mathematical description of neural ODE
dynamics on event manifolds. We demonstrate its versatility across diverse
applications: characterising uncertainties in a data-driven prey-predator
control model, analysing neural optimal feedback dynamics, and mapping landing
trajectories in a three-body neural Hamiltonian system. In all cases, our
method enhances the interpretability and rigour of neural ODEs by expressing
their behaviour through explicit mathematical structures. Our findings
contribute to a deeper theoretical foundation for event-triggered neural
differential equations and provide a mathematical construct for explaining
complex system dynamics.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 08:57:34 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Izzo",
"Dario",
""
],
[
"Origer",
"Sebastien",
""
],
[
"Acciarini",
"Giacomo",
""
],
[
"Biscani",
"Francesco",
""
]
] |
2504.08770 | Joshua Owolabi Adeleke | Joshua Owolabi Adeleke | Symmetry Analysis of Proper-Time Maxwell's Equations: Invariant
Solutions and Physical Implications | null | null | null | null | physics.class-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Gill and Zachary's proper-time reformulation of Maxwell's equations
introduces a source-dependent speed b = sqrt(c^2 + u^2), distinct from standard
electrodynamics. This study applies Lie symmetry analysis to the resulting wave
equations, yielding new invariant solutions, conservation laws, and physical
interpretations. In the one-dimensional case, we derive E(x, tau) = A x / tau +
B + k * integral of rho(x) dx, revealing radiative behavior linked to b.
Conservation laws via self-adjointness, singularity analysis near tau = 0, and
simulations of accelerating charges extend Gill and Zachary's framework. The
findings suggest new insights into propagation and energy dynamics, clarified
as coordinate effects rather than violations of causality, offering fresh
perspectives on relativistic electrodynamics.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 18:18:02 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Adeleke",
"Joshua Owolabi",
""
]
] |
2504.08771 | Chenghao Zhang | Chao Feng, Yanze Zhang, Chenghao Zhang | Generate the browsing process for short-video recommendation | null | null | null | null | cs.IR cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces a new model to generate the browsing process for
short-video recommendation and proposes a novel Segment Content Aware Model via
User Engagement Feedback (SCAM) for watch time prediction in video
recommendation. Unlike existing methods that rely on multimodal features for
video content understanding, SCAM implicitly models video content through
users' historical watching behavior, enabling segment-level understanding
without complex multimodal data. By dividing videos into segments based on
duration and employing a Transformer-like architecture, SCAM captures the
sequential dependence between segments while mitigating duration bias.
Extensive experiments on industrial-scale and public datasets demonstrate
SCAM's state-of-the-art performance in watch time prediction. The proposed
approach offers a scalable and effective solution for video recommendation by
leveraging segment-level modeling and users' engagement feedback.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2025 20:54:52 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Feng",
"Chao",
""
],
[
"Zhang",
"Yanze",
""
],
[
"Zhang",
"Chenghao",
""
]
] |
2504.08772 | Younghwan Lee | Younghwan Lee, Tung M. Luu, Donghoon Lee, and Chang D. Yoo | Reward Generation via Large Vision-Language Model in Offline
Reinforcement Learning | 5 pages, ICASSP 2025. First two authors are equally contributed | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | In offline reinforcement learning (RL), learning from fixed datasets presents
a promising solution for domains where real-time interaction with the
environment is expensive or risky. However, designing dense reward signals for
offline dataset requires significant human effort and domain expertise.
Reinforcement learning with human feedback (RLHF) has emerged as an
alternative, but it remains costly due to the human-in-the-loop process,
prompting interest in automated reward generation models. To address this, we
propose Reward Generation via Large Vision-Language Models (RG-VLM), which
leverages the reasoning capabilities of LVLMs to generate rewards from offline
data without human involvement. RG-VLM improves generalization in long-horizon
tasks and can be seamlessly integrated with the sparse reward signals to
enhance task performance, demonstrating its potential as an auxiliary reward
signal.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 07:11:18 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lee",
"Younghwan",
""
],
[
"Luu",
"Tung M.",
""
],
[
"Lee",
"Donghoon",
""
],
[
"Yoo",
"Chang D.",
""
]
] |
2504.08773 | Olivier Jeunen | Olivier Jeunen | Counterfactual Inference under Thompson Sampling | null | null | null | null | cs.IR cs.LG stat.ME | http://creativecommons.org/licenses/by/4.0/ | Recommender systems exemplify sequential decision-making under uncertainty,
strategically deciding what content to serve to users, to optimise a range of
potential objectives. To balance the explore-exploit trade-off successfully,
Thompson sampling provides a natural and widespread paradigm to
probabilistically select which action to take. Questions of causal and
counterfactual inference, which underpin use-cases like offline evaluation, are
not straightforward to answer in these contexts. Specifically, whilst most
existing estimators rely on action propensities, these are not readily
available under Thompson sampling procedures.
We derive exact and efficiently computable expressions for action
propensities under a variety of parameter and outcome distributions, enabling
the use of off-policy estimators in Thompson sampling scenarios. This opens up
a range of practical use-cases where counterfactual inference is crucial,
including unbiased offline evaluation of recommender systems, as well as
general applications of causal inference in online advertising,
personalisation, and beyond.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 14:31:40 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Jeunen",
"Olivier",
""
]
] |
2504.08774 | Austin Blitstein | Austin M. Blitstein, Rodolfo R. Rosales, Pedro J. S\'aenz | Anomalous interference drives oscillatory dynamics in wave-dressed
active particles | 6 pages, 5 figures | null | null | null | cond-mat.soft physics.flu-dyn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A recent surge of discoveries has sparked significant interest in active
systems where a particle moves autonomously due to a resonant interaction with
its self-generated wave field, leading to notable wave-mediated effects
including new propulsion mechanisms, spontaneous oscillatory dynamics, and
quantum-like phenomena. Drawing from an archetypical model of wave-dressed
active particles, we unveil a wave-mediated non-local force driving their
dynamics, arising from the particle's path memory and an unconventional form of
wave interference near jerking points, locations where the particle's velocity
changes rapidly. In contrast to the typical case of constructive interference
at points of stationary phase, waves excited by the particle near jerking
points avoid cancellation through rapid changes in frequency. Through an
asymptotic analysis, we derive the wave force from jerking points, revealing it
as an elusive but crucial remnant of the particle's past motion that allows us
to rationalize mechanistically in-line speed oscillations, wave-like statistics
in potential wells, and non-specular reflections. The results we derive follow
from generic wave superposition principles, suggesting their applicability to a
broad class of wave-dressed active particles.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 15:42:38 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Blitstein",
"Austin M.",
""
],
[
"Rosales",
"Rodolfo R.",
""
],
[
"Sáenz",
"Pedro J.",
""
]
] |
2504.08775 | Christopher Wolfram | Christopher Wolfram, Aaron Schein | Layers at Similar Depths Generate Similar Activations Across LLM
Architectures | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | How do the latent spaces used by independently-trained LLMs relate to one
another? We study the nearest neighbor relationships induced by activations at
different layers of 24 open-weight LLMs, and find that they 1) tend to vary
from layer to layer within a model, and 2) are approximately shared between
corresponding layers of different models. Claim 2 shows that these nearest
neighbor relationships are not arbitrary, as they are shared across models, but
Claim 1 shows that they are not "obvious" either, as there is no single set of
nearest neighbor relationships that is universally shared. Together, these
suggest that LLMs generate a progression of activation geometries from layer to
layer, but that this entire progression is largely shared between models,
stretched and squeezed to fit into different architectures.
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 21:02:30 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Wolfram",
"Christopher",
""
],
[
"Schein",
"Aaron",
""
]
] |
2504.08776 | Gautam Kishore Shahi | Gautam Kishore Shahi, Oshani Seneviratne, Marc Spaniol | SemCAFE: When Named Entities make the Difference Assessing Web Source
Reliability through Entity-level Analytics | null | null | null | null | cs.CL cs.CY cs.LG | http://creativecommons.org/licenses/by/4.0/ | With the shift from traditional to digital media, the online landscape now
hosts not only reliable news articles but also a significant amount of
unreliable content. Digital media has faster reachability by significantly
influencing public opinion and advancing political agendas. While newspaper
readers may be familiar with their preferred outlets political leanings or
credibility, determining unreliable news articles is much more challenging. The
credibility of many online sources is often opaque, with AI generated content
being easily disseminated at minimal cost. Unreliable news articles,
particularly those that followed the Russian invasion of Ukraine in 2022,
closely mimic the topics and writing styles of credible sources, making them
difficult to distinguish. To address this, we introduce SemCAFE, a system
designed to detect news reliability by incorporating entity relatedness into
its assessment. SemCAFE employs standard Natural Language Processing
techniques, such as boilerplate removal and tokenization, alongside entity
level semantic analysis using the YAGO knowledge base. By creating a semantic
fingerprint for each news article, SemCAFE could assess the credibility of
46,020 reliable and 3,407 unreliable articles on the 2022 Russian invasion of
Ukraine. Our approach improved the macro F1 score by 12% over state of the art
methods. The sample data and code are available on GitHub
| [
{
"version": "v1",
"created": "Thu, 3 Apr 2025 22:14:43 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Shahi",
"Gautam Kishore",
""
],
[
"Seneviratne",
"Oshani",
""
],
[
"Spaniol",
"Marc",
""
]
] |
2504.08777 | Teo Susnjak | Teo Susnjak, Cole Palffy, Tatiana Zimina, Nazgul Altynbekova, Kunal
Garg, Leona Gilbert | The Lyme Disease Controversy: An AI-Driven Discourse Analysis of a
Quarter Century of Academic Debate and Divides | null | null | null | null | cs.CY cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The scientific discourse surrounding Chronic Lyme Disease (CLD) and
Post-Treatment Lyme Disease Syndrome (PTLDS) has evolved over the past
twenty-five years into a complex and polarised debate, shaped by shifting
research priorities, institutional influences, and competing explanatory
models. This study presents the first large-scale, systematic examination of
this discourse using an innovative hybrid AI-driven methodology, combining
large language models with structured human validation to analyse thousands of
scholarly abstracts spanning 25 years. By integrating Large Language Models
(LLMs) with expert oversight, we developed a quantitative framework for
tracking epistemic shifts in contested medical fields, with applications to
other content analysis domains. Our analysis revealed a progressive transition
from infection-based models of Lyme disease to immune-mediated explanations for
persistent symptoms. This study offers new empirical insights into the
structural and epistemic forces shaping Lyme disease research, providing a
scalable and replicable methodology for analysing discourse, while underscoring
the value of AI-assisted methodologies in social science and medical research.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 01:28:29 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Susnjak",
"Teo",
""
],
[
"Palffy",
"Cole",
""
],
[
"Zimina",
"Tatiana",
""
],
[
"Altynbekova",
"Nazgul",
""
],
[
"Garg",
"Kunal",
""
],
[
"Gilbert",
"Leona",
""
]
] |
2504.08778 | Bo Xiong | Bo Xiong, Steffen Staab | From Tokens to Lattices: Emergent Lattice Structures in Language Models | ICLR 2025 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Pretrained masked language models (MLMs) have demonstrated an impressive
capability to comprehend and encode conceptual knowledge, revealing a lattice
structure among concepts. This raises a critical question: how does this
conceptualization emerge from MLM pretraining? In this paper, we explore this
problem from the perspective of Formal Concept Analysis (FCA), a mathematical
framework that derives concept lattices from the observations of
object-attribute relationships. We show that the MLM's objective implicitly
learns a \emph{formal context} that describes objects, attributes, and their
dependencies, which enables the reconstruction of a concept lattice through
FCA. We propose a novel framework for concept lattice construction from
pretrained MLMs and investigate the origin of the inductive biases of MLMs in
lattice structure learning. Our framework differs from previous work because it
does not rely on human-defined concepts and allows for discovering "latent"
concepts that extend beyond human definitions. We create three datasets for
evaluation, and the empirical results verify our hypothesis.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 04:28:17 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Xiong",
"Bo",
""
],
[
"Staab",
"Steffen",
""
]
] |
2504.08779 | Ruoxin Xiong | Ruoxin Xiong, Yanyu Wang, Suat Gunhan, Yimin Zhu, and Charles Berryman | Can AI Master Construction Management (CM)? Benchmarking
State-of-the-Art Large Language Models on CM Certification Exams | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The growing complexity of construction management (CM) projects, coupled with
challenges such as strict regulatory requirements and labor shortages, requires
specialized analytical tools that streamline project workflow and enhance
performance. Although large language models (LLMs) have demonstrated
exceptional performance in general reasoning tasks, their effectiveness in
tackling CM-specific challenges, such as precise quantitative analysis and
regulatory interpretation, remains inadequately explored. To bridge this gap,
this study introduces CMExamSet, a comprehensive benchmarking dataset
comprising 689 authentic multiple-choice questions sourced from four nationally
accredited CM certification exams. Our zero-shot evaluation assesses overall
accuracy, subject areas (e.g., construction safety), reasoning complexity
(single-step and multi-step), and question formats (text-only,
figure-referenced, and table-referenced). The results indicate that GPT-4o and
Claude 3.7 surpass typical human pass thresholds (70%), with average accuracies
of 82% and 83%, respectively. Additionally, both models performed better on
single-step tasks, with accuracies of 85.7% (GPT-4o) and 86.7% (Claude 3.7).
Multi-step tasks were more challenging, reducing performance to 76.5% and
77.6%, respectively. Furthermore, both LLMs show significant limitations on
figure-referenced questions, with accuracies dropping to approximately 40%. Our
error pattern analysis further reveals that conceptual misunderstandings are
the most common (44.4% and 47.9%), underscoring the need for enhanced
domain-specific reasoning models. These findings underscore the potential of
LLMs as valuable supplementary analytical tools in CM, while highlighting the
need for domain-specific refinements and sustained human oversight in complex
decision making.
| [
{
"version": "v1",
"created": "Fri, 4 Apr 2025 18:13:45 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Xiong",
"Ruoxin",
""
],
[
"Wang",
"Yanyu",
""
],
[
"Gunhan",
"Suat",
""
],
[
"Zhu",
"Yimin",
""
],
[
"Berryman",
"Charles",
""
]
] |
2504.08781 | Xu-Xiang Zhong | Xu-Xiang Zhong and Chao Yi and Han-Jia Ye | Efficient Evaluation of Large Language Models via Collaborative
Filtering | null | null | null | null | cs.CL cs.AI cs.IR | http://creativecommons.org/licenses/by/4.0/ | With the development of Large Language Models (LLMs), numerous benchmarks
have been proposed to measure and compare the capabilities of different LLMs.
However, evaluating LLMs is costly due to the large number of test instances
and their slow inference speed. In this paper, we aim to explore how to
efficiently estimate a model's real performance on a given benchmark based on
its evaluation results on a small number of instances sampled from the
benchmark. Inspired by Collaborative Filtering (CF) in Recommendation Systems
(RS), we treat LLMs as users and test instances as items and propose a
two-stage method. In the first stage, we treat instance selection as
recommending products to users to choose instances that can easily distinguish
model performance. In the second stage, we see performance prediction as rating
prediction problem in RS to predict the target LLM's behavior on unselected
instances. Experiments on multiple LLMs and datasets imply that our method can
accurately estimate the target model's performance while largely reducing its
inference overhead.
| [
{
"version": "v1",
"created": "Sat, 5 Apr 2025 07:46:30 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhong",
"Xu-Xiang",
""
],
[
"Yi",
"Chao",
""
],
[
"Ye",
"Han-Jia",
""
]
] |
2504.08782 | Lucas Beerens | Lucas Beerens, Desmond J. Higham | Embedding Hidden Adversarial Capabilities in Pre-Trained Diffusion
Models | null | null | null | null | cs.LG cs.AI cs.CR | http://creativecommons.org/licenses/by/4.0/ | We introduce a new attack paradigm that embeds hidden adversarial
capabilities directly into diffusion models via fine-tuning, without altering
their observable behavior or requiring modifications during inference. Unlike
prior approaches that target specific images or adjust the generation process
to produce adversarial outputs, our method integrates adversarial functionality
into the model itself. The resulting tampered model generates high-quality
images indistinguishable from those of the original, yet these images cause
misclassification in downstream classifiers at a high rate. The
misclassification can be targeted to specific output classes. Users can employ
this compromised model unaware of its embedded adversarial nature, as it
functions identically to a standard diffusion model. We demonstrate the
effectiveness and stealthiness of our approach, uncovering a covert attack
vector that raises new security concerns. These findings expose a risk arising
from the use of externally-supplied models and highlight the urgent need for
robust model verification and defense mechanisms against hidden threats in
generative models. The code is available at
https://github.com/LucasBeerens/CRAFTed-Diffusion .
| [
{
"version": "v1",
"created": "Sat, 5 Apr 2025 12:51:36 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Beerens",
"Lucas",
""
],
[
"Higham",
"Desmond J.",
""
]
] |
2504.08783 | Eduardo Fraga Lima De Melo | Eduardo Fraga L. de Melo and Rodrigo S. Targino | Mensura\c{c}\~ao da Transfer\^encia de Riqueza em Planos de
Contribui\c{c}\~ao Definida com a Marca\c{c}\~ao de Ativos na Curva | in Portuguese language | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The methodology for measuring financial assets in defined contribution (DC)
pension plans has significant implications whether wealth transfers will occur
among participants. In December 2024, a regulatory act was issued for Closed
Pension Entities, allowing the use of the hold-to-maturity (HTM) measurement
method of treasury bonds in DC plans. This article quantifies the financial
impact on participants of adopting HTM valuation in these plans, using real
data from the term structure of the real interest rates to assess the resulting
wealth transfers. The analysis highlights how HTM valuation creates asymmetries
in financial outcomes, benefiting some participants at the expense of others.
Wealth transfers occur both during any withdrawal of funds and at the time of
contributions, including portfolio reallocations that involve buying or selling
bonds. Partial use of HTM or attempts to immunize outflows do not completely
eliminate wealth transfers. The results reinforce that the use of
mark-to-market (MTM) valuation of assets in DC plans prevents wealth transfers
and, consequently, financial losses for participants. O m\'etodo de
mensura\c{c}\~ao de ativos financeiros em planos de previd\^encia na modalidade
de contribui\c{c}\~ao definida (CD, ou contribui\c{c}\~ao vari\'avel CV, na
fase de acumula\c{c}\~ao) tem implica\c{c}\~oes significativas se haver\'a
transfer\^encia de riqueza entre os participantes. Em Dez/2024 foi publicada
norma para as Entidades Fechadas de Previd\^encia Complementar possibilitando o
uso da marca\c{c}\~ao na curva de t\'itulos p\'ublicos federais nos planos CD e
CV na fase de acumula\c{c}\~ao. Este artigo quantifica o impacto financeiro nos
participantes da ado\c{c}\~ao da marca\c{c}\~ao na curva (HTM {\it Hold to
Maturity}) nestes planos, utilizando dados reais da estrutura a termo da taxa
de juros de cupom de IPCA para avaliar as transfer\^encias de riqueza
resultantes dessa ado\c{c}\~ao. A an\'alise evidencia como a marca\c{c}\~ao na
curva gera assimetrias nos resultados financeiros, beneficiando alguns
participantes em detrimento de outros. As transfer\^encias de riqueza ocorrem
tanto em qualquer retirada de recursos quanto tamb\'em na entrada
(contribui\c{c}\~oes), inclusive realoca\c{c}\~oes da carteira que impliquem
venda ou compra de t\'itulos. O uso do HTM de forma parcial ou a tentativa de
imuniza\c{c}\~ao de sa\'idas n\~ao eliminam por completo transfer\^encias de
riqueza. Os resultados refor\c{c}am que, para fins de cotiza\c{c}\~ao, o uso da
marca\c{c}\~ao a mercado (MTM {\it Mark to Market}) de ativos em planos CD (e
CV na fase de diferimento) evita transfer\^encias de riqueza e, por
consequ\^encia, preju\'izos financeiros aos seus participantes.
| [
{
"version": "v1",
"created": "Sat, 5 Apr 2025 15:01:32 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"de Melo",
"Eduardo Fraga L.",
""
],
[
"Targino",
"Rodrigo S.",
""
]
] |
2504.08784 | Siyuan Chen | Siyuan Chen, Zhipeng Jia, Samira Khan, Arvind Krishnamurthy, Phillip
B. Gibbons | SLOs-Serve: Optimized Serving of Multi-SLO LLMs | null | null | null | null | cs.DC cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | This paper introduces SLOs-Serve, a system designed for serving multi-stage
large language model (LLM) requests with application- and stage-specific
service level objectives (SLOs). The key idea behind SLOs-Serve is to customize
the allocation of tokens to meet these SLO requirements. SLOs-Serve uses a
multi-SLO dynamic programming-based algorithm to continuously optimize token
allocations under SLO constraints by exploring the full design space of chunked
prefill and (optional) speculative decoding. Leveraging this resource planning
algorithm, SLOs-Serve effectively supports multi-SLOs and multi-replica serving
with dynamic request routing while being resilient to bursty arrivals. Our
evaluation across 6 LLM application scenarios (including summarization, coding,
chatbot, tool calling, and reasoning) demonstrates that SLOs-Serve improves
per-GPU serving capacity by 2.2x on average compared to prior state-of-the-art
systems.
| [
{
"version": "v1",
"created": "Sat, 5 Apr 2025 17:41:26 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Chen",
"Siyuan",
""
],
[
"Jia",
"Zhipeng",
""
],
[
"Khan",
"Samira",
""
],
[
"Krishnamurthy",
"Arvind",
""
],
[
"Gibbons",
"Phillip B.",
""
]
] |
2504.08786 | Tong Zhang | Tong Zhang | AdaptRec: A Self-Adaptive Framework for Sequential Recommendations with
Large Language Models | null | null | null | null | cs.IR cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The recent advancements in Large Language Models (LLMs) have generated
considerable interest in their utilization for sequential recommendation tasks.
While collaborative signals from similar users are central to recommendation
modeling, effectively transforming these signals into a format that LLMs can
understand and utilize remains challenging. The critical challenges include
selecting relevant demonstrations from large-scale user interactions and
ensuring their alignment with LLMs' reasoning process. To address these
challenges, we introduce AdaptRec, a self-adaptive fram-ework that leverages
LLMs for sequential recommendations by incorporating explicit collaborative
signals. AdaptRec employs a two-phase user selection mechanism -- User
Similarity Retrieval and Self-Adaptive User Selection -- to efficiently
identify relevant user sequences in large-scale datasets from multi-metric
evaluation. We also develop a User-Based Similarity Retrieval Prompt, enabling
the model to actively select similar users and continuously refine its
selection criteria during training. Using the collaborative signals from
similar users, we construct a User-Contextualized Recommendation Prompt that
translates their behavior sequences into natural language, explicitly
integrating this information into the recommendation process. Experiments
demonstrate AdaptRec's superior performance, with significant improvements in
HitRatio@1 scores of 7.13\%, 18.16\%, and 10.41\% across real-world datasets
with full fine-tuning, and even higher gains of 23.00\%, 15.97\%, and 17.98\%
in few-shot scenarios.
| [
{
"version": "v1",
"created": "Sun, 6 Apr 2025 00:30:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Tong",
""
]
] |
2504.08787 | Wan Shou | Youwen Liang, Wan Shou | Molecular Dynamics (MD) simulation of silicon nanoparticle
crystallization during laser-induced forward transfer (LIFT) printing | null | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Laser-induced forward transfer (LIFT) printing is a versatile technique to
realize micro/nano-scale additive manufacturing of functional materials,
including metals, and semiconductors. However, the crystallization phenomena
during LIFT printing have not been well understood. In this work, we attempt to
gain a comprehensive understanding of silicon crystallization during LIFT
printing. Specifically, molecular dynamics (MD) simulation is used to
investigate the size effect on the melting and crystallization of Si
nanoparticles during transportation in air. We found with the decrease in
nanoparticle size, crystallization becomes rare, even with a low cooling rate.
The nucleation location of different particles is also analyzed and almost
always starts at a sub-surface location (below 5 {\AA}). The atomic structure
evolution during solidification is also monitored to provide guidance for LIFT
printing of Si. Our simulation results indicate that nano-confinement can lead
to single-crystal structure formation, which may shed light on single-crystal
additive manufacturing.
| [
{
"version": "v1",
"created": "Sun, 6 Apr 2025 01:43:07 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Liang",
"Youwen",
""
],
[
"Shou",
"Wan",
""
]
] |
2504.08789 | Abdullah Guvendi | Semra Gurtas Dogan, Kobra Hasanirokh, Omar Mustafa and Abdullah
Guvendi | Twist-Induced Effects on Weyl Pairs in Magnetized Graphene Nanoribbons | 7 pages, 8 figures | null | null | null | cond-mat.mes-hall cond-mat.mtrl-sci hep-th quant-ph | http://creativecommons.org/licenses/by/4.0/ | This paper presents an analytical investigation into the dynamics of Weyl
pairs within magnetized helicoidal graphene nanoribbons. By embedding a curved
surface into flat Minkowski space-time, we derive a fully covariant two-body
Dirac equation specific to this system. We begin by formulating a
non-perturbative wave equation that governs the relative motion of the Weyl
pairs and obtain exact solutions. Our results demonstrate the influence of the
uniform magnetic field and the number of twists on the dynamics of Weyl pairs
in graphene nanoribbons, providing precise energy values that lay a robust
foundation for future research. Furthermore, we examine the material's response
to perturbation fields by calculating the polarization function and
investigating how twisting and magnetic fields affect this response. Our
findings indicate that, in principle, the material's properties, which are
crucial for practical applications, can be effectively controlled by precisely
tuning the magnetic field and the number of twists in graphene nanoribbons.
| [
{
"version": "v1",
"created": "Sun, 6 Apr 2025 11:25:24 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Dogan",
"Semra Gurtas",
""
],
[
"Hasanirokh",
"Kobra",
""
],
[
"Mustafa",
"Omar",
""
],
[
"Guvendi",
"Abdullah",
""
]
] |
2504.08790 | Ejaz Ahmed | Ejaz Ahmed (1), Huajie Zhang (1), Mert Aybar (1), Bikai Jin (2),
Shihao Wang (2) and Rui Qiao (1) ((1) Department of Mechanical Engineering,
Virginia Tech, Blacksburg, USA, (2) Chevron Energy Technology Co., Houston,
USA) | Imbibition of Oil in Dry and Prewetted Calcite Nanopores | 24 pages, 8 figures, Submitted to Physics of Fluids, Rui Qiao: To
whom correspondence should be addressed. Email: [email protected] | null | null | null | cond-mat.soft physics.atom-ph physics.flu-dyn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fluid imbibition into porous media featuring nanopores is ubiquitous in
applications such as oil recovery from unconventional reservoirs and material
processing. While the imbibition of pure fluids has been extensively studied,
the imbibition of fluid mixture is little explored. Here we report the
molecular dynamics study of the imbibition of model crude oil into
nanometer-wide mineral pores, both when pore walls are dry and prewetted by a
residual water film. Results show the fastest imbibition and quickest
propagation of molecularly thin precursor films ahead of the oil meniscus in
the dry pore system. The presence of a thin water film on pore walls
corresponding to an environmental relative humidity of 30% slows down but still
allows the spontaneous imbibition of single-component oil. Introducing polar
components into the oil slows down the imbibition into dry nanopores, due
partly to the clogging of the pore entrance. Strong selectivity toward nonpolar
oil is evident. The slowdown of imbibition by polar oil is less significant in
the prewetted pores than in dry pores, but the selectivity toward nonpolar oil
remains strong.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 00:30:30 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ahmed",
"Ejaz",
""
],
[
"Zhang",
"Huajie",
""
],
[
"Aybar",
"Mert",
""
],
[
"Jin",
"Bikai",
""
],
[
"Wang",
"Shihao",
""
],
[
"Qiao",
"Rui",
""
]
] |
2504.08791 | Zonghang Li | Zonghang Li, Tao Li, Wenjiao Feng, Mohsen Guizani, Hongfang Yu | PRIMA.CPP: Speeding Up 70B-Scale LLM Inference on Low-Resource Everyday
Home Clusters | 23 pages, 9 figures, 6 tables | null | null | null | cs.DC cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Emergency of DeepSeek R1 and QwQ 32B have broken through performance barriers
for running frontier large language models (LLMs) on home devices. While
consumer hardware is getting stronger and model quantization is improving,
existing end-side solutions still demand GPU clusters, large RAM/VRAM, and high
bandwidth, far beyond what a common home cluster can handle. This paper
introduces prima.cpp, a distributed inference system that runs 70B-scale models
on everyday home devices using a mix of CPU/GPU, low RAM/VRAM, Wi-Fi, and
cross-platform support. It uses mmap to manage model weights and introduces
piped-ring parallelism with prefetching to hide disk loading. By modeling
heterogeneity in computation, communication, disk, memory (and its management
behavior), and OS, it optimally assigns model layers to each device's CPU and
GPU, further reducing token latency. An elegant algorithm named Halda is
proposed to solve this NP-hard assignment problem. We evaluate prima.cpp on a
common four-node home cluster. It outperforms llama.cpp, exo, and dllama on
30B+ models while keeping memory pressure below 6%. This brings frontier
30B-70B models, such as Llama 3, DeepSeek R1, Qwen 2.5, and QwQ to home
assistants, making advanced AI truly accessible to individuals. The code is
open source and available at https://github.com/Lizonghang/prima.cpp.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 13:46:21 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Li",
"Zonghang",
""
],
[
"Li",
"Tao",
""
],
[
"Feng",
"Wenjiao",
""
],
[
"Guizani",
"Mohsen",
""
],
[
"Yu",
"Hongfang",
""
]
] |
2504.08792 | Toqeer Ehsan | Toqeer Ehsan and Thamar Solorio | Enhancing NER Performance in Low-Resource Pakistani Languages using
Cross-Lingual Data Augmentation | Accepted to W-NUT 2025 @ NAACL | null | null | null | cs.CL cs.IR | http://creativecommons.org/licenses/by-sa/4.0/ | Named Entity Recognition (NER), a fundamental task in Natural Language
Processing (NLP), has shown significant advancements for high-resource
languages. However, due to a lack of annotated datasets and limited
representation in Pre-trained Language Models (PLMs), it remains understudied
and challenging for low-resource languages. To address these challenges, we
propose a data augmentation technique that generates culturally plausible
sentences and experiments on four low-resource Pakistani languages; Urdu,
Shahmukhi, Sindhi, and Pashto. By fine-tuning multilingual masked Large
Language Models (LLMs), our approach demonstrates significant improvements in
NER performance for Shahmukhi and Pashto. We further explore the capability of
generative LLMs for NER and data augmentation using few-shot learning.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 15:18:34 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ehsan",
"Toqeer",
""
],
[
"Solorio",
"Thamar",
""
]
] |
2504.08793 | Jorge A Huertas | Jorge A. Huertas and Pascal Van Hentenryck | A Constraint Programming Model For Serial Batch Scheduling With Minimum
Batch Size | 13 pages, 7 figures | null | null | null | cs.DC cs.AI math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In serial batch (s-batch) scheduling, jobs are grouped in batches and
processed sequentially within their batch. This paper considers multiple
parallel machines, nonidentical job weights and release times, and
sequence-dependent setup times between batches of different families. Although
s-batch has been widely studied in the literature, very few papers have taken
into account a minimum batch size, typical in practical settings such as
semiconductor manufacturing and the metal industry. The problem with this
minimum batch size requirement has been mostly tackled with dynamic programming
and meta-heuristics, and no article has ever used constraint programming (CP)
to do so. This paper fills this gap by proposing, for the first time, a CP
model for s-batching with minimum batch size. The computational experiments on
standard cases compare the CP model with two existing mixed-integer programming
(MIP) models from the literature. The results demonstrate the versatility of
the proposed CP model to handle multiple variations of s-batching; and its
ability to produce, in large instances, better solutions than the MIP models
faster.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 17:14:19 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Huertas",
"Jorge A.",
""
],
[
"Van Hentenryck",
"Pascal",
""
]
] |
2504.08794 | Emilio Mart\'inez-Pa\~neda | T. Hageman, C. Andrade, E. Mart\'inez-Pa\~neda | Corrosion of metal reinforcements within concrete and localisation of
supporting reactions under natural conditions | null | null | null | null | physics.chem-ph cond-mat.mtrl-sci cs.CE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Corrosion in concrete prevents in-situ observation, necessitating models to
provide insight into the local reaction currents. We present a computational
method for predicting corrosion rates of reinforcements within concrete under
natural conditions, i.e. requiring the corrosion current to be supported by
equal cathodic currents. In contrast to typical corrosion models, where these
two counteracting currents are required to be co-located, we allow these
currents to be separated such that pitting corrosion can be supported by
cathodic reactions over a much larger area. Pitting corrosion is investigated,
elucidating the effects of the concrete porosity and water saturation, the
presence of dissolved oxygen, and chlorine concentration within the pore
solution. The presented model is capable of capturing the dynamic growth of
acidic regions around corrosion pits, showing the limited region over which the
hydrogen evolution reaction occurs and how this region evolves over time. The
ability of oxygen to diffuse towards the metal surface due to increased
porosity is seen to have a major effect on the corrosion rate, whereas changes
in the chlorine concentration (and thus changes in the conductivity of the pore
solution) play a secondary role. Furthermore, external oxygen is seen to
enhance corrosion but is not required to initialise and sustain acidic
corrosion pits.
| [
{
"version": "v1",
"created": "Mon, 7 Apr 2025 18:21:58 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Hageman",
"T.",
""
],
[
"Andrade",
"C.",
""
],
[
"Martínez-Pañeda",
"E.",
""
]
] |
2504.08795 | Amir Fakhim Babaei | Amir Fakhim Babaei, Thidapat Chantem | DARIS: An Oversubscribed Spatio-Temporal Scheduler for Real-Time DNN
Inference on GPUs | null | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The widespread use of Deep Neural Networks (DNNs) is limited by high
computational demands, especially in constrained environments. GPUs, though
effective accelerators, often face underutilization and rely on coarse-grained
scheduling. This paper introduces DARIS, a priority-based real-time DNN
scheduler for GPUs, utilizing NVIDIA's MPS and CUDA streaming for spatial
sharing, and a synchronization-based staging method for temporal partitioning.
In particular, DARIS improves GPU utilization and uniquely analyzes GPU
concurrency by oversubscribing computing resources. It also supports zero-delay
DNN migration between GPU partitions. Experiments show DARIS improves
throughput by 15% and 11.5% over batching and state-of-the-art schedulers,
respectively, even without batching. All high-priority tasks meet deadlines,
with low-priority tasks having under 2% deadline miss rate. High-priority
response times are 33% better than those of low-priority tasks.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 01:13:38 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Babaei",
"Amir Fakhim",
""
],
[
"Chantem",
"Thidapat",
""
]
] |
2504.08796 | Yao-guang Zheng | Winston Chen, Yao-Guang Zheng | Advanced Analysis of Hawking Temperature Calculation for Novel
Topological Black Holes using Laurent Series and the RVB Method | null | null | null | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper employs Laurent series expansions and the
Robson--Villari--Biancalana (RVB) method to provide a refined derivation of the
Hawking temperature for two newly introduced topological black hole solutions.
Previous calculations have demonstrated inconsistencies when applying
traditional methods to such exotic horizons, prompting the need for a more
thorough mathematical analysis. By systematically incorporating higher-order
terms in the Laurent expansions of the metric functions near the horizon and
leveraging the topological features characterized by the Euler characteristic,
we reveal additional corrections to the Hawking temperature beyond standard
approaches. These findings underscore the subtle interplay between local
geometry, spacetime topology, and quantum effects. The results clarify
discrepancies found in earlier works, present a more accurate representation of
thermodynamic properties for the black holes in question, and suggest broader
implications for topological structures in advanced gravitational theories.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 04:04:22 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Chen",
"Winston",
""
],
[
"Zheng",
"Yao-Guang",
""
]
] |
2504.08797 | Federico Re | Federico Re and Oliver F. Piattella | Non zero Coriolis field in Ehlers' Frame Theory | 15 pages, no figures | Galaxies, 13(2), 38 (2025) | 10.3390/galaxies13020038 | null | gr-qc astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | Ehlers' Frame Theory is a class of geometric theories parameterized by
$\lambda := 1/c^2$ and identical to the General Theory of Relativity for
$\lambda \neq 0$. The limit $\lambda \to 0$ does not recover Newtonian gravity,
as one might expect, but yields the so-called Newton-Cartan theory of gravity,
which is characterized by a second gravitational field $\boldsymbol{\omega}$,
called the Coriolis field. Such a field encodes at a non-relativistic level the
dragging feature of general spacetimes, as we show explicitly for the case of
the $(\eta,H)$ geometries. Taking advantage of the Coriolis field, we apply
Ehlers' theory to an axially symmetric distribution of matter, mimicking, for
example, a disc galaxy, and show how its dynamics might reproduce a flattish
rotation curve. In the same setting, we further exploit the formal simplicity
of Ehlers' formalism in addressing non-stationary cases, which are remarkably
difficult to be treated in the General Theory of Relativity. We show that the
time derivative of the Coriolis field gives rise to a tangential acceleration
which allows to study a possible formation in time of the rotation curve's
flattish feature.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 09:33:04 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Re",
"Federico",
""
],
[
"Piattella",
"Oliver F.",
""
]
] |
2504.08798 | Leo Yu Zhang Dr. | Xiaomei Zhang, Zhaoxi Zhang, Yanjun Zhang, Xufei Zheng, Leo Yu Zhang,
Shengshan Hu, and Shirui Pan | Exploring Gradient-Guided Masked Language Model to Detect Textual
Adversarial Attacks | null | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Textual adversarial examples pose serious threats to the reliability of
natural language processing systems. Recent studies suggest that adversarial
examples tend to deviate from the underlying manifold of normal texts, whereas
pre-trained masked language models can approximate the manifold of normal data.
These findings inspire the exploration of masked language models for detecting
textual adversarial attacks. We first introduce Masked Language Model-based
Detection (MLMD), leveraging the mask and unmask operations of the masked
language modeling (MLM) objective to induce the difference in manifold changes
between normal and adversarial texts. Although MLMD achieves competitive
detection performance, its exhaustive one-by-one masking strategy introduces
significant computational overhead. Our posterior analysis reveals that a
significant number of non-keywords in the input are not important for detection
but consume resources. Building on this, we introduce Gradient-guided MLMD
(GradMLMD), which leverages gradient information to identify and skip
non-keywords during detection, significantly reducing resource consumption
without compromising detection performance.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 14:10:57 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Xiaomei",
""
],
[
"Zhang",
"Zhaoxi",
""
],
[
"Zhang",
"Yanjun",
""
],
[
"Zheng",
"Xufei",
""
],
[
"Zhang",
"Leo Yu",
""
],
[
"Hu",
"Shengshan",
""
],
[
"Pan",
"Shirui",
""
]
] |
2504.08799 | Francisco Bento Lustosa | C\'elio R. Muniz, Francisco Bento Lustosa and Takol Tangphati | Einstein-Yang-Mills Regular Black Holes in Rainbow Gravity | null | null | null | null | gr-qc | http://creativecommons.org/licenses/by/4.0/ | In this work, we investigate regular black hole solutions in nonminimal
Einstein-Yang-Mills theory modified by Rainbow Gravity, focusing on the impact
of quantum gravity effects on their thermodynamics, particle emission, energy
conditions, curvature, and shadow formation. We find that the rainbow parameter
$\lambda$ alters Hawking's temperature, entropy, and specific heat, leading to
modified phase transitions and the possible formation of remnants. We calculate
the graybody factor demonstrating that particle emission is enhanced with
increasing $\lambda$, reflecting the behavior of the temperature and confirming
the impact of the rainbow parameter on the evaporation process. Energy
conditions are violated inside the black hole, with violations intensifying for
larger $\lambda$. We also show that Rainbow Gravity mitigates singularity
formation by softening the curvature near the origin, contributing to the
regularity of the solution. Finally, we study the black hole shadow and
demonstrate that its radius decreases as quantum gravity effects strengthen,
suggesting potential observational tests for Rainbow Gravity. These results
highlight the role of Rainbow Gravity in modifying black hole physics and
provides a framework for exploring quantum gravitational corrections in
astrophysical scenarios.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 15:34:58 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Muniz",
"Célio R.",
""
],
[
"Lustosa",
"Francisco Bento",
""
],
[
"Tangphati",
"Takol",
""
]
] |
2504.08800 | Gaurav Gadbail | Gaurav N. Gadbail | Accelerated Expansion of the Universe in Nonmetricity-based Modified
Gravity | Ph.D. Thesis (113 Pages). Chapters are published in the following
journals: Physics Letters B 835, 137509 (2022); Astrophysical Journal 972,
174 (2024); Physics Letters B 860, 139232 (2024); Fortschritte der Physik 73,
2400205 (2025) | null | null | null | gr-qc | http://creativecommons.org/licenses/by/4.0/ | This thesis explores the cosmological implications of modified gravity,
focusing on nonmetricity-based $f(Q)$ gravity as an alternative to the
$\Lambda$CDM model in explaining cosmic acceleration. Chapter I lays the
theoretical groundwork by reviewing General Relativity (GR), $\Lambda$CDM, and
the limitations of the standard model, motivating $f(Q)$ gravity. Chapter II
constructs cosmological reconstructions of $f(Q)$ gravity within the FLRW
framework, deriving forms of $f(Q)$ that replicate the $\Lambda$CDM expansion
and using the e-folding parameter to show compatibility with various cosmic
histories. Chapter III addresses challenges with arbitrary $f(Q)$ forms by
applying Gaussian Process (GP) reconstruction using observational Hubble data.
This model-independent method reconstructs the Hubble parameter H(z), leading
to a data-driven $f(Q)$ form. Motivated by this, a new parametrization $f(Q) =
-2\Lambda + \epsilon Q^2$ is proposed, and power-law and exponential models are
tested for consistency. Chapter IV incorporates a quintessence scalar field in
power-law $f(Q)$ gravity to study inflation and late-time acceleration. Using
GP, the scalar potential $V(\phi)$ is reconstructed and analyzed. Results show
that early dark energy has little impact today, but reconstructed quintessence
models offer insights into cosmic acceleration. Chapter V examines interacting
dark energy and matter under power-law $f(Q)$ using dynamical systems. Two
interaction types are studied, and fixed points linked to de Sitter and
quintessence solutions are identified. Chapter VI concludes and suggests future
work.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 19:05:08 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Gadbail",
"Gaurav N.",
""
]
] |
2504.08801 | Andrew Kiruluta | Andrew Kiruluta, Priscilla Burity and Samantha Williams | Learnable Multi-Scale Wavelet Transformer: A Novel Alternative to
Self-Attention | null | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by/4.0/ | Transformer architectures, underpinned by the self-attention mechanism, have
achieved state-of-the-art results across numerous natural language processing
(NLP) tasks by effectively modeling long-range dependencies. However, the
computational complexity of self-attention, scaling quadratically with input
sequence length, presents significant challenges for processing very long
sequences or operating under resource constraints. This paper introduces the
Learnable Multi-Scale Wavelet Transformer (LMWT), a novel architecture that
replaces the standard dot-product self-attention with a learnable multi-scale
Haar wavelet transform module. Leveraging the intrinsic multi-resolution
properties of wavelets, the LMWT efficiently captures both local details and
global context. Crucially, the parameters of the wavelet transform, including
scale-specific coefficients, are learned end-to-end during training, allowing
the model to adapt its decomposition strategy to the data and task. We present
the detailed mathematical formulation of the learnable Haar wavelet module and
its integration into the transformer framework, supplemented by an
architectural diagram. We conduct a comprehensive experimental evaluation on a
standard machine translation benchmark (WMT16 En-De), comparing the LMWT
against a baseline self-attention transformer using metrics like BLEU score,
perplexity, and token accuracy. Furthermore, we analyze the computational
complexity, highlighting the linear scaling of our approach, discuss its
novelty in the context of related work, and explore the interpretability
offered by visualizing the learned Haar coefficients. Our results indicate that
the LMWT achieves competitive performance while offering substantial
computational advantages, positioning it as a promising and novel alternative
for efficient sequence modeling.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 22:16:54 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Kiruluta",
"Andrew",
""
],
[
"Burity",
"Priscilla",
""
],
[
"Williams",
"Samantha",
""
]
] |
2504.08802 | David Johnson | David R. Johnson, Smita Krishnaswamy, Michael Perlmutter | InfoGain Wavelets: Furthering the Design of Diffusion Wavelets for
Graph-Structured Data | This work was accepted to be presented at the Graph Signal Processing
Workshop 2025 | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion wavelets extract information from graph signals at different scales
of resolution by utilizing graph diffusion operators raised to various powers,
known as diffusion scales. Traditionally, the diffusion scales are chosen to be
dyadic integers, $\mathbf{2^j}$. Here, we propose a novel, unsupervised method
for selecting the diffusion scales based on ideas from information theory. We
then show that our method can be incorporated into wavelet-based GNNs via graph
classification experiments.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 23:13:29 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Johnson",
"David R.",
""
],
[
"Krishnaswamy",
"Smita",
""
],
[
"Perlmutter",
"Michael",
""
]
] |
2504.08803 | Zezhi Tang | Zezhi Tang, Xiaoyu Chen, Xin Jin, Benyuan Zhang, Wenyu Liang | A temporal scale transformer framework for precise remaining useful life
prediction in fuel cells | 7 figs, 10 pages | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In exploring Predictive Health Management (PHM) strategies for Proton
Exchange Membrane Fuel Cells (PEMFC), the Transformer model, widely used in
data-driven approaches, excels in many fields but struggles with time series
analysis due to its self-attention mechanism, which yields a complexity of the
input sequence squared and low computational efficiency. It also faces
challenges in capturing both global long-term dependencies and local details
effectively. To tackle this, we propose the Temporal Scale Transformer
(TSTransformer), an enhanced version of the inverted Transformer
(iTransformer). Unlike traditional Transformers that treat each timestep as an
input token, TSTransformer maps sequences of varying lengths into tokens at
different stages for inter-sequence modeling, using attention to capture
multivariate correlations and feed-forward networks (FFN) to encode sequence
representations. By integrating a one-dimensional convolutional layer into the
multivariate attention for multi-level scaling of K and V matrices, it improves
local feature extraction, captures temporal scale characteristics, and reduces
token count and computational costs. Experiments comparing TSTransformer with
models like Long Short-Term Memory, iTransformer, and Transformer demonstrate
its potential as a powerful tool for advancing PHM in renewable energy,
effectively addressing the limitations of pure Transformer models in
data-driven time series tasks.
| [
{
"version": "v1",
"created": "Tue, 8 Apr 2025 23:42:54 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Tang",
"Zezhi",
""
],
[
"Chen",
"Xiaoyu",
""
],
[
"Jin",
"Xin",
""
],
[
"Zhang",
"Benyuan",
""
],
[
"Liang",
"Wenyu",
""
]
] |
2504.08804 | Pooya Razavi | Pooya Razavi, Sonya J. Powers | Estimating Item Difficulty Using Large Language Models and Tree-Based
Machine Learning Algorithms | null | null | null | null | cs.CY cs.CL cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Estimating item difficulty through field-testing is often resource-intensive
and time-consuming. As such, there is strong motivation to develop methods that
can predict item difficulty at scale using only the item content. Large
Language Models (LLMs) represent a new frontier for this goal. The present
research examines the feasibility of using an LLM to predict item difficulty
for K-5 mathematics and reading assessment items (N = 5170). Two estimation
approaches were implemented: (a) a direct estimation method that prompted the
LLM to assign a single difficulty rating to each item, and (b) a feature-based
strategy where the LLM extracted multiple cognitive and linguistic features,
which were then used in ensemble tree-based models (random forests and gradient
boosting) to predict difficulty. Overall, direct LLM estimates showed moderate
to strong correlations with true item difficulties. However, their accuracy
varied by grade level, often performing worse for early grades. In contrast,
the feature-based method yielded stronger predictive accuracy, with
correlations as high as r = 0.87 and lower error estimates compared to both
direct LLM predictions and baseline regressors. These findings highlight the
promise of LLMs in streamlining item development and reducing reliance on
extensive field testing and underscore the importance of structured feature
extraction. We provide a seven-step workflow for testing professionals who
would want to implement a similar item difficulty estimation approach with
their item pool.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 00:04:07 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Razavi",
"Pooya",
""
],
[
"Powers",
"Sonya J.",
""
]
] |
2504.08805 | Kleanthis Karakolios | James Bono, Justin Grana, Kleanthis Karakolios, Pruthvi Hanumanthapura
Ramakrishna, and Ankit Srivastava | Generative AI in Live Operations: Evidence of Productivity Gains in
Cybersecurity and Endpoint Management | null | null | null | null | cs.CR cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We measure the association between generative AI (GAI) tool adoption and four
metrics spanning security operations, information protection, and endpoint
management: 1) number of security alerts per incident, 2) probability of
security incident reopenings, 3) time to classify a data loss prevention alert,
and 4) time to resolve device policy conflicts. We find that GAI is associated
with robust and statistically and practically significant improvements in the
four metrics. Although unobserved confounders inhibit causal identification,
these results are among the first to use observational data from live
operations to investigate the relationship between GAI adoption and security
operations, data loss prevention, and device policy management.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 00:34:20 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Bono",
"James",
""
],
[
"Grana",
"Justin",
""
],
[
"Karakolios",
"Kleanthis",
""
],
[
"Ramakrishna",
"Pruthvi Hanumanthapura",
""
],
[
"Srivastava",
"Ankit",
""
]
] |
2504.08806 | Ling Luo | Luo Ling, Bai Qianqian | Endowing Embodied Agents with Spatial Reasoning Capabilities for
Vision-and-Language Navigation | null | null | null | null | cs.AI cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Enhancing the spatial perception capabilities of mobile robots is crucial for
achieving embodied Vision-and-Language Navigation (VLN). Although significant
progress has been made in simulated environments, directly transferring these
capabilities to real-world scenarios often results in severe hallucination
phenomena, causing robots to lose effective spatial awareness. To address this
issue, we propose BrainNav, a bio-inspired spatial cognitive navigation
framework inspired by biological spatial cognition theories and cognitive map
theory. BrainNav integrates dual-map (coordinate map and topological map) and
dual-orientation (relative orientation and absolute orientation) strategies,
enabling real-time navigation through dynamic scene capture and path planning.
Its five core modules-Hippocampal Memory Hub, Visual Cortex Perception Engine,
Parietal Spatial Constructor, Prefrontal Decision Center, and Cerebellar Motion
Execution Unit-mimic biological cognitive functions to reduce spatial
hallucinations and enhance adaptability. Validated in a zero-shot real-world
lab environment using the Limo Pro robot, BrainNav, compatible with GPT-4,
outperforms existing State-of-the-Art (SOTA) Vision-and-Language Navigation in
Continuous Environments (VLN-CE) methods without fine-tuning.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 02:19:22 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Ling",
"Luo",
""
],
[
"Qianqian",
"Bai",
""
]
] |
2504.08807 | Zhiyue Wu | HongZheng Liu and YiNuo Tian and Zhiyue Wu | The Exploratory Study on the Relationship Between the Failure of
Distance Metrics in High-Dimensional Space and Emergent Phenomena | null | null | null | null | cs.IT cond-mat.stat-mech math.IT nlin.AO | http://creativecommons.org/licenses/by/4.0/ | This paper presents a unified framework, integrating information theory and
statistical mechanics, to connect metric failure in high-dimensional data with
emergence in complex systems. We propose the "Information Dilution Theorem,"
demonstrating that as dimensionality ($d$) increases, the mutual information
efficiency between geometric metrics (e.g., Euclidean distance) and system
states decays approximately as $O(1/d)$. This decay arises from the mismatch
between linearly growing system entropy and sublinearly growing metric entropy,
explaining the mechanism behind distance concentration. Building on this, we
introduce information structural complexity ($C(S)$) based on the mutual
information matrix spectrum and interaction encoding capacity ($C'$) derived
from information bottleneck theory. The "Emergence Critical Theorem" states
that when $C(S)$ exceeds $C'$, new global features inevitably emerge,
satisfying a predefined mutual information threshold. This provides an
operational criterion for self-organization and phase transitions. We discuss
potential applications in physics, biology, and deep learning, suggesting
potential directions like MI-based manifold learning (UMAP+) and offering a
quantitative foundation for analyzing emergence across disciplines.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 02:19:58 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Liu",
"HongZheng",
""
],
[
"Tian",
"YiNuo",
""
],
[
"Wu",
"Zhiyue",
""
]
] |
2504.08808 | Zhengke Sun | Zhengke Sun, Hangwei Qian, Ivor Tsang | Exploring the Effectiveness and Interpretability of Texts in LLM-based
Time Series Models | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) have been applied to time series forecasting
tasks, leveraging pre-trained language models as the backbone and incorporating
textual data to purportedly enhance the comprehensive capabilities of LLMs for
time series. However, are these texts really helpful for interpretation? This
study seeks to investigate the actual efficacy and interpretability of such
textual incorporations. Through a series of empirical experiments on textual
prompts and textual prototypes, our findings reveal that the misalignment
between two modalities exists, and the textual information does not
significantly improve time series forecasting performance in many cases.
Furthermore, visualization analysis indicates that the textual representations
learned by existing frameworks lack sufficient interpretability when applied to
time series data. We further propose a novel metric named Semantic Matching
Index (SMI) to better evaluate the matching degree between time series and
texts during our post hoc interpretability investigation. Our analysis reveals
the misalignment and limited interpretability of texts in current time-series
LLMs, and we hope this study can raise awareness of the interpretability of
texts for time series. The code is available at
https://github.com/zachysun/TS-Lang-Exp.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 02:48:35 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Sun",
"Zhengke",
""
],
[
"Qian",
"Hangwei",
""
],
[
"Tsang",
"Ivor",
""
]
] |
2504.08809 | Wei Chen | Wei Chen, Xin Yan, Bin Wen, Fan Yang, Tingting Gao, Di Zhang, Long
Chen | Decoupling Contrastive Decoding: Robust Hallucination Mitigation in
Multimodal Large Language Models | 13 pages, 4 figures | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Although multimodal large language models (MLLMs) exhibit remarkable
reasoning capabilities on complex multimodal understanding tasks, they still
suffer from the notorious hallucination issue: generating outputs misaligned
with obvious visual or factual evidence. Currently, training-based solutions,
like direct preference optimization (DPO), leverage paired preference data to
suppress hallucinations. However, they risk sacrificing general reasoning
capabilities due to the likelihood displacement. Meanwhile, training-free
solutions, like contrastive decoding, achieve this goal by subtracting the
estimated hallucination pattern from a distorted input. Yet, these handcrafted
perturbations (e.g., add noise to images) may poorly capture authentic
hallucination patterns. To avoid these weaknesses of existing methods, and
realize robust hallucination mitigation (i.e., maintaining general reasoning
performance), we propose a novel framework: Decoupling Contrastive Decoding
(DCD). Specifically, DCD decouples the learning of positive and negative
samples in preference datasets, and trains separate positive and negative image
projections within the MLLM. The negative projection implicitly models real
hallucination patterns, which enables vision-aware negative images in the
contrastive decoding inference stage. Our DCD alleviates likelihood
displacement by avoiding pairwise optimization and generalizes robustly without
handcrafted degradation. Extensive ablations across hallucination benchmarks
and general reasoning tasks demonstrate the effectiveness of DCD, i.e., it
matches DPO's hallucination suppression while preserving general capabilities
and outperforms the handcrafted contrastive decoding methods.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 02:59:18 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Chen",
"Wei",
""
],
[
"Yan",
"Xin",
""
],
[
"Wen",
"Bin",
""
],
[
"Yang",
"Fan",
""
],
[
"Gao",
"Tingting",
""
],
[
"Zhang",
"Di",
""
],
[
"Chen",
"Long",
""
]
] |
2504.08810 | Yingming Pu | Zheyuan Lai and Yingming Pu | PriM: Principle-Inspired Material Discovery through Multi-Agent
Collaboration | null | null | null | null | cs.LG cond-mat.mtrl-sci cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Complex chemical space and limited knowledge scope with biases holds immense
challenge for human scientists, yet in automated materials discovery. Existing
intelligent methods relies more on numerical computation, leading to
inefficient exploration and results with hard-interpretability. To bridge this
gap, we introduce a principles-guided material discovery system powered by
language inferential multi-agent system (MAS), namely PriM. Our framework
integrates automated hypothesis generation with experimental validation in a
roundtable system of MAS, enabling systematic exploration while maintaining
scientific rigor. Based on our framework, the case study of nano helix
demonstrates higher materials exploration rate and property value while
providing transparent reasoning pathways. This approach develops an
automated-and-transparent paradigm for material discovery, with broad
implications for rational design of functional materials. Code is publicly
available at our \href{https://github.com/amair-lab/PriM}{GitHub}.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 03:05:10 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Lai",
"Zheyuan",
""
],
[
"Pu",
"Yingming",
""
]
] |
2504.08811 | Zirui Chen | Zirui Chen, Zhaoyang Zhang, Ziqing Xing, Ridong Li, Zhaohui Yang,
Richeng Jin, Chongwen Huang, Yuzhi Yang and M\'erouane Debbah | Analogical Learning for Cross-Scenario Generalization: Framework and
Application to Intelligent Localization | null | null | null | null | cs.LG cs.CE eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing learning models often exhibit poor generalization when deployed
across diverse scenarios. It is mainly due to that the underlying reference
frame of the data varies with the deployment environment and settings. However,
despite the data of each scenario has its distinct reference frame, its
generation generally follows the same underlying physical rule. Based on these
findings, this article proposes a brand-new universal deep learning framework
named analogical learning (AL), which provides a highly efficient way to
implicitly retrieve the reference frame information associated with a scenario
and then to make accurate prediction by relative analogy across scenarios.
Specifically, an elegant bipartite neural network architecture called
Mateformer is designed, the first part of which calculates the relativity
within multiple feature spaces between the input data and a small amount of
embedded data from the current scenario, while the second part uses these
relativity to guide the nonlinear analogy. We apply AL to the typical
multi-scenario learning problem of intelligent wireless localization in
cellular networks. Extensive experiments show that AL achieves state-of-the-art
accuracy, stable transferability and robust adaptation to new scenarios without
any tuning, and outperforming conventional methods with a precision improvement
of nearly two orders of magnitude. All data and code are available at
https://github.com/ziruichen-research/ALLoc.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 03:36:52 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Chen",
"Zirui",
""
],
[
"Zhang",
"Zhaoyang",
""
],
[
"Xing",
"Ziqing",
""
],
[
"Li",
"Ridong",
""
],
[
"Yang",
"Zhaohui",
""
],
[
"Jin",
"Richeng",
""
],
[
"Huang",
"Chongwen",
""
],
[
"Yang",
"Yuzhi",
""
],
[
"Debbah",
"Mérouane",
""
]
] |
2504.08812 | David Johnston | David O. Johnston, Arkajyoti Chakraborty, Nora Belrose | Mechanistic Anomaly Detection for "Quirky" Language Models | ICLR Building Trust Workshop 2025 | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by/4.0/ | As LLMs grow in capability, the task of supervising LLMs becomes more
challenging. Supervision failures can occur if LLMs are sensitive to factors
that supervisors are unaware of. We investigate Mechanistic Anomaly Detection
(MAD) as a technique to augment supervision of capable models; we use internal
model features to identify anomalous training signals so they can be
investigated or discarded. We train detectors to flag points from the test
environment that differ substantially from the training environment, and
experiment with a large variety of detector features and scoring rules to
detect anomalies in a set of ``quirky'' language models. We find that detectors
can achieve high discrimination on some tasks, but no detector is effective
across all models and tasks. MAD techniques may be effective in low-stakes
applications, but advances in both detection and evaluation are likely needed
if they are to be used in high stakes settings.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 06:03:18 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Johnston",
"David O.",
""
],
[
"Chakraborty",
"Arkajyoti",
""
],
[
"Belrose",
"Nora",
""
]
] |
2504.08813 | Junfeng Fang | Junfeng Fang, Yukai Wang, Ruipeng Wang, Zijun Yao, Kun Wang, An Zhang,
Xiang Wang, Tat-Seng Chua | SafeMLRM: Demystifying Safety in Multi-modal Large Reasoning Models | null | null | null | null | cs.LG cs.AI cs.CR | http://creativecommons.org/licenses/by/4.0/ | The rapid advancement of multi-modal large reasoning models (MLRMs) --
enhanced versions of multimodal language models (MLLMs) equipped with reasoning
capabilities -- has revolutionized diverse applications. However, their safety
implications remain underexplored. While prior work has exposed critical
vulnerabilities in unimodal reasoning models, MLRMs introduce distinct risks
from cross-modal reasoning pathways. This work presents the first systematic
safety analysis of MLRMs through large-scale empirical studies comparing MLRMs
with their base MLLMs. Our experiments reveal three critical findings: (1) The
Reasoning Tax: Acquiring reasoning capabilities catastrophically degrades
inherited safety alignment. MLRMs exhibit 37.44% higher jailbreaking success
rates than base MLLMs under adversarial attacks. (2) Safety Blind Spots: While
safety degradation is pervasive, certain scenarios (e.g., Illegal Activity)
suffer 25 times higher attack rates -- far exceeding the average 3.4 times
increase, revealing scenario-specific vulnerabilities with alarming cross-model
and datasets consistency. (3) Emergent Self-Correction: Despite tight
reasoning-answer safety coupling, MLRMs demonstrate nascent self-correction --
16.9% of jailbroken reasoning steps are overridden by safe answers, hinting at
intrinsic safeguards. These findings underscore the urgency of scenario-aware
safety auditing and mechanisms to amplify MLRMs' self-correction potential. To
catalyze research, we open-source OpenSafeMLRM, the first toolkit for MLRM
safety evaluation, providing unified interface for mainstream models, datasets,
and jailbreaking methods. Our work calls for immediate efforts to harden
reasoning-augmented AI, ensuring its transformative potential aligns with
ethical safeguards.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 06:53:23 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Fang",
"Junfeng",
""
],
[
"Wang",
"Yukai",
""
],
[
"Wang",
"Ruipeng",
""
],
[
"Yao",
"Zijun",
""
],
[
"Wang",
"Kun",
""
],
[
"Zhang",
"An",
""
],
[
"Wang",
"Xiang",
""
],
[
"Chua",
"Tat-Seng",
""
]
] |
2504.08814 | Aakar Mathur | Aakar Mathur, Ashish Gupta, Sajal K. Das | When Federated Learning Meets Quantum Computing: Survey and Research
Opportunities | submitted to IEEE Communications Surveys and Tutorials | null | null | null | cs.DC cs.ET cs.LG cs.NE | http://creativecommons.org/licenses/by/4.0/ | Quantum Federated Learning (QFL) is an emerging field that harnesses advances
in Quantum Computing (QC) to improve the scalability and efficiency of
decentralized Federated Learning (FL) models. This paper provides a systematic
and comprehensive survey of the emerging problems and solutions when FL meets
QC, from research protocol to a novel taxonomy, particularly focusing on both
quantum and federated limitations, such as their architectures, Noisy
Intermediate Scale Quantum (NISQ) devices, and privacy preservation, so on.
This work explores key developments and integration strategies, along with the
impact of quantum computing on FL, keeping a sharp focus on hybrid
quantum-classical approaches. The paper offers an in-depth understanding of how
the strengths of QC, such as gradient hiding, state entanglement, quantum key
distribution, quantum security, and quantum-enhanced differential privacy, have
been integrated into FL to ensure the privacy of participants in an enhanced,
fast, and secure framework. Finally, this study proposes potential future
directions to address the identified research gaps and challenges, aiming to
inspire faster and more secure QFL models for practical use.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 07:29:33 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Mathur",
"Aakar",
""
],
[
"Gupta",
"Ashish",
""
],
[
"Das",
"Sajal K.",
""
]
] |
2504.08815 | Jayabrata Dhar Dr. | Samarendu Biswas, Sachidananda Mahato, and Jayabrata Dhar | Bioconvection and Bloom in Porous environment and Its Assisted
Technologies: A Review | null | null | null | null | physics.bio-ph | http://creativecommons.org/licenses/by/4.0/ | Bioconvection, a phenomenon arising from the collective motion of motile
microorganisms, plays a crucial role in shaping microbial distributions, fluid
dynamics, and bloom formation in aquatic environments. While extensive research
has explored bioconvection in open waters, a significant proportion of
microbial life exists within porous habitats, including soils, sediments, and
subsurface environments, where the interplay between microbial motility and
structural heterogeneity remains less understood. This review synthesizes
recent advancements in the study of bioconvective dynamics within porous media,
emphasizing the impacts of various external parameters on microbial transport,
and spatial organization in structured environments. Additionally, we explore
the various scenarios where bloom formation interacts with porous ecosystems.
We further discuss the implications of microbial interactions with porous
substrates for bioremediation, health management, and industrial applications.
A dedicated section explores genetic engineering approaches aimed at modulating
microbial motility and growth characteristics to enhance bioconvective
behaviour and bloom regulation. By integrating experimental, numerical, and
theoretical insights, this review highlights key knowledge gaps and outlines
future research directions toward a deeper understanding of microbial transport
and bioconvection in porous systems, with potential applications in
environmental sustainability and biotechnology.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 07:40:58 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Biswas",
"Samarendu",
""
],
[
"Mahato",
"Sachidananda",
""
],
[
"Dhar",
"Jayabrata",
""
]
] |
2504.08816 | Sicheng Liu | Sicheng Liu, Hongchang Huang, Bo Yang, Mingxuan Cai, Xu Yang, Xinping
Guan | A Graph-Enhanced DeepONet Approach for Real-Time Estimating
Hydrogen-Enriched Natural Gas Flow under Variable Operations | null | null | null | null | cs.LG cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Blending green hydrogen into natural gas presents a promising approach for
renewable energy integration and fuel decarbonization. Accurate estimation of
hydrogen fraction in hydrogen-enriched natural gas (HENG) pipeline networks is
crucial for operational safety and efficiency, yet it remains challenging due
to complex dynamics. While existing data-driven approaches adopt end-to-end
architectures for HENG flow state estimation, their limited adaptability to
varying operational conditions hinders practical applications. To this end,
this study proposes a graph-enhanced DeepONet framework for the real-time
estimation of HENG flow, especially hydrogen fractions. First, a dual-network
architecture, called branch network and trunk network, is employed to
characterize operational conditions and sparse sensor measurements to estimate
the HENG state at targeted locations and time points. Second, a graph-enhance
branch network is proposed to incorporate pipeline topology, improving the
estimation accuracy in large-scale pipeline networks. Experimental results
demonstrate that the proposed method achieves superior estimation accuracy for
HCNG flow under varying operational conditions compared to conventional
approaches.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 09:42:50 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Liu",
"Sicheng",
""
],
[
"Huang",
"Hongchang",
""
],
[
"Yang",
"Bo",
""
],
[
"Cai",
"Mingxuan",
""
],
[
"Yang",
"Xu",
""
],
[
"Guan",
"Xinping",
""
]
] |
2504.08817 | Takahiro Misawa | Takahiro Misawa, Ai Koizumi, Ryo Tamura, and Kazuyoshi Yoshimi | Exploring utilization of generative AI for research and education in
data-driven materials science | 13 pages, 3 figures | null | null | null | cs.CY cs.AI cs.LG physics.ed-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generative AI has recently had a profound impact on various fields, including
daily life, research, and education. To explore its efficient utilization in
data-driven materials science, we organized a hackathon -- AIMHack2024 -- in
July 2024. In this hackathon, researchers from fields such as materials
science, information science, bioinformatics, and condensed matter physics
worked together to explore how generative AI can facilitate research and
education. Based on the results of the hackathon, this paper presents topics
related to (1) conducting AI-assisted software trials, (2) building AI tutors
for software, and (3) developing GUI applications for software. While
generative AI continues to evolve rapidly, this paper provides an early record
of its application in data-driven materials science and highlights strategies
for integrating AI into research and education.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 11:15:21 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Misawa",
"Takahiro",
""
],
[
"Koizumi",
"Ai",
""
],
[
"Tamura",
"Ryo",
""
],
[
"Yoshimi",
"Kazuyoshi",
""
]
] |
2504.08818 | Xinyu Zhang | Xinyu Zhang, Shanshan Feng, Xutao Li | From Text to Time? Rethinking the Effectiveness of the Large Language
Model for Time Series Forecasting | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Using pre-trained large language models (LLMs) as the backbone for time
series prediction has recently gained significant research interest. However,
the effectiveness of LLM backbones in this domain remains a topic of debate.
Based on thorough empirical analyses, we observe that training and testing
LLM-based models on small datasets often leads to the Encoder and Decoder
becoming overly adapted to the dataset, thereby obscuring the true predictive
capabilities of the LLM backbone. To investigate the genuine potential of LLMs
in time series prediction, we introduce three pre-training models with
identical architectures but different pre-training strategies. Thereby,
large-scale pre-training allows us to create unbiased Encoder and Decoder
components tailored to the LLM backbone. Through controlled experiments, we
evaluate the zero-shot and few-shot prediction performance of the LLM, offering
insights into its capabilities. Extensive experiments reveal that although the
LLM backbone demonstrates some promise, its forecasting performance is limited.
Our source code is publicly available in the anonymous repository:
https://anonymous.4open.science/r/LLM4TS-0B5C.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 13:20:09 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Zhang",
"Xinyu",
""
],
[
"Feng",
"Shanshan",
""
],
[
"Li",
"Xutao",
""
]
] |
2504.08819 | Dorota W\k{e}g{\l}owska Dr | Rafa{\l} W\k{e}g{\l}owski, Anna Spad{\l}o, Dorota W\k{e}g{\l}owska | Banana DNA derivatives as homeotropic alignment layers in optical
devices | This work was supported by the National Centre of Science MINIATURA 6
2022/06/X/ST5/00508 and UGB 22-720 and 22723. The authors thank Mateusz
Mrukiewicz for performing ionic conductivity measurements | Soft Matter, 2024,20, 8561-8569 | 10.1039/D4SM00322E | null | cond-mat.soft cond-mat.mtrl-sci physics.bio-ph | http://creativecommons.org/licenses/by/4.0/ | In this study, deoxyribonucleic acid (DNA) from bananas was extracted and
functionalized and used for the first time as a homeotropic alignment layer for
liquid crystals (LCs). Our research was aimed at extracting and investigating
DNA from bananas via the synthesis and study of DNA complexes with various
surfactants to examine the usefulness of such a complex as an alignment layer
in electro-optical transducers. We proposed a simple and eco-friendly synthesis
of the DNA complexes isolated from bananas with surfactants, so we transformed
the DNA isolated from bananas into a functionalized alignment layer. A
biopolymer alignment layer like deoxyribonucleic acid (DNA) from a banana
complexed with a cationic surfactant is an excellent alternative to a commonly
used but toxic polyimide alignment layer. DNA-based materials are promising for
photonic applications and biosensors because of their excellent optical and
physical properties, biodegradability, and low production cost. The novelty of
the research lies in the potential use of these materials as biodegradable
biopolymer alignment layers for optical devices instead of conventional
polymers, which are usually harmful for the environment.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 13:27:42 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Węgłowski",
"Rafał",
""
],
[
"Spadło",
"Anna",
""
],
[
"Węgłowska",
"Dorota",
""
]
] |
2504.08820 | Jing Yao | Jing Yao, Xiaoyuan Yi, Jindong Wang, Zhicheng Dou, Xing Xie | CAReDiO: Cultural Alignment of LLM via Representativeness and
Distinctiveness Guided Data Optimization | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As Large Language Models (LLMs) more deeply integrate into human life across
various regions, aligning them with pluralistic cultures is crucial for
improving user experience and mitigating cultural conflicts. Existing
approaches develop culturally aligned LLMs primarily through fine-tuning with
massive carefully curated culture-specific corpora. Nevertheless, inspired by
culture theories, we identify two key challenges faced by these datasets: (1)
Representativeness: These corpora fail to fully capture the target culture's
core characteristics with redundancy, causing computation waste; (2)
Distinctiveness: They struggle to distinguish the unique nuances of a given
culture from shared patterns across other relevant ones, hindering precise
cultural modeling. To handle these challenges, we introduce CAReDiO, a novel
cultural data construction framework. Specifically, CAReDiO utilizes powerful
LLMs to automatically generate cultural conversation data, where both the
queries and responses are further optimized by maximizing representativeness
and distinctiveness. Using CAReDiO, we construct a small yet effective dataset,
covering five cultures, and compare it with several recent cultural corpora.
Extensive experiments demonstrate that our method generates more effective data
and enables cultural alignment with as few as 100 training samples, enhancing
both performance and efficiency.
| [
{
"version": "v1",
"created": "Wed, 9 Apr 2025 13:40:13 GMT"
}
] | 2025-04-15T00:00:00 | [
[
"Yao",
"Jing",
""
],
[
"Yi",
"Xiaoyuan",
""
],
[
"Wang",
"Jindong",
""
],
[
"Dou",
"Zhicheng",
""
],
[
"Xie",
"Xing",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.