| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| SelfCheckGPT | MEMIT | Reliability Score62,500,000,000 | 64 | 1mo ago | |
| TruthfulQA gated mode (test) | DoLa | IGR (%)77.9 | 28 | 2mo ago | |
| WikiBigEdit | GRACE | Error Rate (ERR)1 | 24 | 4mo ago | |
| UniEdit | GRACE | Error Rate (ERR)1 | 24 | 4mo ago | |
| Hallucination | SoLA | Error Rate (ERR)15.15 | 10 | 4mo ago |