| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| MMLU Moral Scenarios (test) | SKIG | Accuracy0.86 | 28 | 3mo ago | |
| Moral OOD | GSPO+AIR | Accuracy81.07 | 13 | 13d ago | |
| Moral ID | GRPO+AIR | Accuracy85.51 | 13 | 13d ago | |
| MoralExceptQA (challenge set) | MORALCOT | F1 Score64.47 | 11 | 3mo ago | |
| UNIMORAL | Qwen2.5-7B-Instruct | Acc (mean)67.9 | 6 | 3mo ago |