| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| Short-context task suite (WikiText, LAMBADA, TriviaQA, PIQA, HellaSwag, WinoGrande, ARC-Easy, GPQA, Social IQA, OpenBookQA, SciQ) (test) | RoPE++EC | WikiText PPL14.4 | 18 | 4d ago | |
| Standard Benchmarks (ARC-E, ARC-C, BoolQ, HellaSwag, OBQA, PIQA, WinoGrande, MMLU, SciQ) (test) | WeSaR-GlobalGC | ARC-E Acc (Norm)49.75 | 8 | 4d ago |