| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| CIFAR-100 (test) | DPS | ECE0.85 | 104 | 4mo ago | |
| HMDA (test) | Oracle ECE1.29 | 100 | 2mo ago | ||
| BEAR (test) | gemma-7b | Brier Score0.083 | 96 | 4mo ago | |
| LFW | AC-Linear | Worst-group Brier score0.051 | 66 | 1mo ago | |
| BFW | FairCal | Worst-Group Brier Score0.042 | 66 | 1mo ago | |
| RFW | AC-Density | Worst-group Brier Score0.051 | 66 | 1mo ago | |
| MMLU | Verbalized confidence | Brier Score0.0559 | 58 | 1mo ago | |
| USPS | ECE1.54 | 57 | 2mo ago | ||
| NQ | Temp. Scaling | ECE0.046 | 55 | 1mo ago | |
| CIFAR-10H | P+L (Recalibrated) | ECE0.84 | 52 | 4mo ago | |
| Brock-Hommes (test) | ANTR | MSE0 | 40 | 4mo ago | |
| TriviaQA | Probe (train on TriviaQA) | Brier Score0.0845 | 39 | 1mo ago | |
| MNIST | ETS | ECE0.21 | 33 | 4mo ago | |
| TruthfulQA | Gain2.132 | 32 | 1mo ago | ||
| SQuAD | Temp. Scaling | ECE5.87 | 31 | 4mo ago | |
| WebQ | Temp. Scaling | ECE0.0674 | 31 | 4mo ago | |
| Digital-S | Knowledge-Transferring-based Temperature Scaling | ECE7.37 | 27 | 4mo ago | |
| Average StrategyQA, HotpotQA, NQ, Bamboogle | NAACL | ECE0.264 | 24 | 4mo ago | |
| Bamboogle | NAACL | ECE0.113 | 24 | 4mo ago | |
| HotpotQA | NAACL | ECE0.28 | 24 | 4mo ago | |
| StrategyQA | NAACL | ECE0.285 | 24 | 4mo ago | |
| ImageNet-C OOD-domains | MD-TS | ECE (%)1.43 | 24 | 4mo ago | |
| ImageNet-C (InD-domains) | TS | ECE (%)0.89 | 24 | 4mo ago | |
| CIFAR-10 5000-sample half (test) | ECE0.0095 | 23 | 2mo ago | ||
| CSQA | CaliDist (Pr.) | ECE1.04 | 22 | 1mo ago |