OpenCompass safety (Perspective toxicity)

OpenCompass runs free-text completions on prompts from a local safety.txt file and scores toxicity with Google Perspective API.

Also known as: safety_gen, safety_datasets

unassessed

This page is a discovery lead. Nobody has yet assessed it against the catalogue contract, so it carries no disposition. Absence of evidence here is not evidence of staleness.

Category	safety
Subcategory	OpenCompass generation-toxicity probe on a local prompt file
Page status	unknown
Metric	Perspective toxicity (expected_max_toxicity, toxic_frac_valid, avg_toxicity_score)
Direction	lower_is_better
Publisher	OpenCompass

What it measures

OpenCompass `safety` is a generation-toxicity probe, not a named standalone paper. The model sees one prompt per line from `./data/safety.txt` and writes a free-text continuation. Perspective API then scores TOXICITY on that continuation. English and Chinese are requested on the API call; the prompt file's language mix is not established here. This is not [helm_safety](helm_safety.md), not [autobencher_safety](autobencher_safety.md), and not the separate OpenCompass RealToxicPrompts config.

Task format

Zero-shot GenInferencer. PromptTemplate is the raw `{prompt}` string. ZeroRetriever. SafetyDataset builds a Hugging Face DatasetDict with a test split only (idx, prompt). output_column is idx because there is no gold string. ToxicEvaluator does not use references.

Models reporting this benchmark

No model card in ModelSpec reports this benchmark yet.

Data

This page as JSON · Edit on GitHub