Copy

src/about.py

@@ -34,18 +34,17 @@ These benchmarks assess fundamental reasoning and knowledge capabilities of mode
 
 | Benchmark           | Description                                                                      |
 |--------------------|----------------------------------------------------------------------------------|
-| **ARC-Easy** / **ARC-Challenge** | Multiple-choice science questions measuring scientific & commonsense reasoning. |
-| **DROP**             | Reading comprehension benchmark emphasizing discrete reasoning steps.             |
-| **WinoGrande**        | Commonsense reasoning challenge focused on co-reference resolution.             |
-| **GSM8K**             | Grade-school math word problems testing arithmetic & multi-step reasoning.         |
-| **HellaSwag**        | Commonsense inference task centered on action completion.                        |
+| **ARC-Easy** / **ARC-Challenge** | Multiple-choice science questions. |
+| **DROP**             | Comprehension benchmark evaluating advanced reasoning capability.             |
+| **WinoGrande**        | Commonsense reasoning challenge.             |
+| **GSM8K**             | Grade-school math word problems testing math capability & multi-step reasoning.         |
+| **HellaSwag**        | Commonsense reasoning task.                        |
 | **HumanEval**         | Evaluates code generation and reasoning in a programming context.                |
-| **IFEval**            | Specialized benchmark for incremental formal reasoning.                          |
-| **IFEval**            | Specialized benchmark for incremental formal reasoning.                          |
-| **MATH**              | High school-level math questions requiring detailed solutions.                  |
+| **IFEval**            | Specialized benchmark for instruction following.                          |
+| **MATH**              | Challenging questions sourced from  math competitions.                |
 | **MMLU** / **MMLU-Pro**| Multi-subject multiple-choice tests of advanced knowledge.                     |
-| **GPQA-Diamond**      | Question-answering benchmark assessing deeper reasoning & knowledge linking.      |
-| **MMMU** (Multi-Choice / Open-Ended) | Multilingual & multi-domain tasks testing structured & open responses.   |
+| **GPQA-Diamond**      | Question-answering benchmark assessing deeper reasoning.      |
+| **MMMU** (Multi-Choice / Open-Ended) | Multi-modal tasks testing structured & open responses.   |
 </div>
 
 ### 🚀 Agentic Benchmarks
@@ -56,11 +55,11 @@ These benchmarks go beyond basic reasoning and evaluate more advanced, autonomou
 
 | Benchmark              | Description                                                                 |
 |-----------------------|----------------------------------------------------------------------------|
-| **GAIA**                | Evaluates autonomous reasoning, planning, problem-solving, & multi-turn interactions. |
-| [**InterCode-CTF**](https://ukgovernmentbeis.github.io/inspect_evals/evals/cybersecurity/in_house_ctf/)   | Capture-the-flag challenge focused on code interpretation & debugging.       |
-| **GDM-In-House-CTF**    | Capture-the-flag challenge testing web application security skills.         |
+| **GAIA**                | Evaluates autonomous reasoning, planning, problem-solving for question answering. |
+| [**InterCode-CTF**]  | Capture-the-flag challenge testing security skills.        |
+| **In-House-CTF**    | Capture-the-flag challenge testing security skills.         |
 | **AgentHarm** / **AgentHarm-Benign** | Measures harmfulness of LLM agents (and benign behavior baseline).   |
-| **SWE-Bench**           | Tests AI agent ability to solve software engineering tasks.                 |
+| **SWE-Bench-Verified**           | Tests AI agent ability to solve software engineering tasks.                 |
 </div>
 """