← Search

Stephen Fitz

2 accepted papers

2024

Safetywashing: Do AI Safety Benchmarks Actually Measure Safety Progress?

NeurIPS 2024poster

Performance on popular ML benchmarks is highly correlated with model scale, suggesting that most benchmarks tend to measure a similar underlying factor of general model capabilities. However, substantial research effort remains devoted to designing new benchmarks, many of which claim to measure nove…

Cited by 22SourcecodeScholar
2024

The WMDP Benchmark: Measuring and Reducing Malicious Use with Unlearning

ICML 2024poster

The White House Executive Order on Artificial Intelligence highlights the risks of large language models (LLMs) empowering malicious actors in developing biological, cyber, and chemical weapons. To measure these risks, government institutions and major AI labs are developing evaluations for hazardou…

Cited by 145SourcePDFScholar