What it measures
Measures explicit and implicit hate speech generation and detection.
Reading the measurement
Separate satisfying instructions from factual accuracy, human preference and safe behavior. Inspect strict all-constraint success as well as average criterion credit. Judge choice and adversarial prompt selection can materially change the result.
Published evaluation
Expected maximum classifier toxicity among 25 sampled continuations. Lower is better; the subscript standard deviation is retained in source-cell notes and excluded from the mean value.
Where this fits in an evaluation suite
Format compliance, preference, truthfulness, and safety require separate measurements. Keep desirable writing style from masking factual errors.
Start with a capability hypothesis, then choose a second test with a different failure mode. Keep the checkpoint, reasoning effort, agent tools, sampling policy, and scoring procedure fixed when measuring a training intervention. Record failures and missing outputs in the denominator.
A reproducible comparison
Expected maximum classifier toxicity among 25 sampled continuations. Lower is better; the subscript standard deviation is retained in source-cell notes and excluded from the mean value.
Publish the exact run configuration, item coverage, score units, and available uncertainty. If a source supplies only an aggregate or a single run, preserve that limit. A new model name or a rounded score alone does not establish a capability improvement.
PRIMARY EVIDENCE & FURTHER READING
Reviewed 2026-10-27. Explanation and interpretation by RLTasks. Benchmark ownership remains with its original authors.