jc-ryan/AgentAlign
Introduction AgentAlign is a comprehensive safety alignment dataset derived from our research work AgentAlign: Navigating Safety Alignment in the Shift from Informative to Agentic Large Language Models. While many models demonstrate robust safety alignment against information-seeking harmful requests (achieving ~90% refusal rates on benchmarks like AdvBench), they show dramatic performance degradation when facing agentic harmful requests, with refusal rates dropping below 20% on… See the full description on the dataset page: https://huggingface.co/datasets/jc-ryan/AgentAlign.
| Author / organization | jc-ryan |
|---|---|
| License (as declared) | apache-2.0 |
| Downloads | 191 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 28 May 2025 |
| Last updated at source | 2 June 2025 |
Popularity, last 2 days
191 now (+4 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.