josancamon/paperbench
PaperBench: Evaluating AI's Ability to Replicate AI Research Paper: arXiv:2504.01848GitHub: openai/preparedness ๐ Dataset Description PaperBench is a benchmark dataset for evaluating the ability of AI agents to replicate state-of-the-art AI research from scratch. The dataset contains 20 ICML 2024 Spotlight and Oral papers, each decomposed into hierarchical rubrics with clear grading criteria. Key Statistics 20 research papers from ICML 2024 8,316โฆ See the full description on the dataset page: https://huggingface.co/datasets/josancamon/paperbench.
| Author / organization | josancamon |
|---|---|
| License (as declared) | mit |
| Downloads | 2.7K (as reported by Hugging Face) |
| Likes | 1 |
| Created | 1 October 2025 |
| Last updated at source | 25 January 2026 |
Popularity, last 3 days
2,661 now (-301 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Recent changes
- 16 Oct 2026 Description: PaperBench: Evaluating AI's Ability to Replicate AI Research Paper: arXiv:2504.01848GitHub: openai/preparedness ๐ Dataset Description PaperBench is a benchmark dataset for evaluating the ability of AI agents to replicate state-of-the-art A โ PaperBench: Evaluating AI's Ability to Replicate AI Research Paper: arXiv:2504.01848GitHub: openai/preparedness ๐ Dataset Description PaperBench is a benchmark dataset for evaluating the ability of AI agents to replicate state-of-the-art A
Changes detected by Silytics between refreshes of the source.
More from josancamon
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.