File size: 1,828 Bytes
d5f0c59 554735b |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 |
---
license: mit
language:
- en
base_model:
- Qwen/Qwen2.5-Coder-7B-Instruct
---
## Introduction
We present a fine-tuned model for formal verification tasks. It is fine-tuned in five formal specification languages (Cog, Dafny, Lean4, ACSL, and TLA) on six formal-verification-related tasks:
- **Requirement Analysis**: given requirements and description of the verification or modeling goals, decomposing the goal into detailed verification steps
- **Proof/Model Generation**: given requirements and description of the verification or modeling goals, writing formal proofs or models that can be verified by verifier/model checker.
- **Proof segment generation**
- **Proof Completion**: complete the given incomplete proofs or models
- **Proof Infilling**: filling in the middle of the given incomplete proofs or models
- **Code 2 Proof**: (Currently only support for ACSL whose specification is in form of code annotations) given the code under verification, generate the proof with the specifications
## Application Scenario
<p align="center">
<img width=100%" src="figures/application.png">
</p>
## Supported Formal Specification Languages
<p align="center">
<img width=100%" src="figures/examples.png">
</p>
## Data Preparation Pipeline
<p align="center">
<img width=60%" src="figures/data-prepare.png">
</p>
## Citation
```
@misc{fmbench25jialun,
title={From Informal to Formal--Incorporating and Evaluating LLMs on Natural Language Requirements to Verifiable Formal Proofs},
author={Jialun Cao, Yaojie Lu, Meiziniu Li, Haoyang Ma, Haokun Li, Mengda He, Cheng Wen, Le Sun, Hongyu Zhang, Shengchao Qin, Shing-Chi Cheung, Cong Tian},
year={2025},
eprint={2501.16207},
archivePrefix={arXiv},
primaryClass={cs.AI},
url={https://arxiv.org/abs/2501.16207},
}
```
|