Center for AI Safety (CAIS) — publication: The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning — benchmark for evaluating dual-use AI capabilities in biosecurity, cybersecurity, and chemical weapons
The source text confirms all key elements of the claim: (1) The publication title matches exactly; (2) It is authored by Li et al. (Nathaniel Li is the first author); (3) Multiple authors are affiliated with Center for AI Safety; (4) The benchmark measures dual-use AI capabilities in biosecurity, cybersecurity, and chemical security; (5) The arxiv ID 2403.03218 matches the source URL, confirming the March 2024 date. The claim accurately represents the paper's content and authorship.
Our claim
entire record- Subject
- Center for AI Safety (CAIS)
- Value
- The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning — benchmark for evaluating dual-use AI capabilities in biosecurity, cybersecurity, and chemical weapons
- As Of
- March 2024
- Notes
- By Li et al.
Source evidence
1 src · 1 checkNoteThe source text confirms all key elements of the claim: (1) The publication title matches exactly; (2) It is authored by Li et al. (Nathaniel Li is the first author); (3) Multiple authors are affiliated with Center for AI Safety; (4) The benchmark measures dual-use AI capabilities in biosecurity, cybersecurity, and chemical security; (5) The arxiv ID 2403.03218 matches the source URL, confirming the March 2024 date. The claim accurately represents the paper's content and authorship.