Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Disallowed Content Evaluation research benchmarks · SOTA2 Research

Disallowed Content Evaluation

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

97.6Non-Overrefusal Rate

4

May 1, 2026

98Not Unsafe Rate

4

May 1, 2026

Challenging Refusal Evaluation

93.4Non-unsafe Rate

4

May 1, 2026

Standard Disallowed Content Evaluation

gpt-5-thinking-mini

99.6Hate (Aggregate)

4

May 1, 2026

Popular datasets

XSTest WildChat Standard Disallowed Content Evaluation Challenging Refusal Evaluation

Follow for updates

Join SOTA2 Contact