DecodingTrust

Benchmark / dataset

Trustworthiness across 8 dimensions (toxicity, bias, privacy, robustness).

Category

Bias, fairness & societal harm

Insurance exposure

Regulatory/bias

Publisher

UIUC / Virtue AI

Publisher type

Academic

License

CC-BY-SA-4.0

Status

Stale (>12mo)

Last commit

2024-09-16

Stars

317

View source on GitHub →

Wanich Risk Lab, 2026.