Share your thoughts, 1 month free Claude Pro on usSee more

Home/Benchmarks

Toxicity Mitigation on Toxic (test)

56Toxicity Score

G

Updated 1mo ago

Evaluation Results

Method	Links
G 2023.07		56	61.5	55.5
PREADD-S 2023.07		43.8	56.5	60

SOTA Paper

G

OPT: Open Pre-trained Transformer Language Models

Dataset

Toxic

Follow for update

@wizwand_team Discord

Related Benchmarks

Toxicity Mitigation on RealToxicityPrompts (test)Toxic Output Mitigation on RealToxicityPrompts 1.0 (Random)Toxic Output Mitigation on RealToxicityPrompts 1.0 (Toxic)Sentiment Control on PosToNeg prompts Sentiment Control on IMDb PosToNeg (test)

© 2026 wizwand

Blog Contact Changelog Swarm

Privacy Policy Terms of Service FAQs Swarm Docs