| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| ValueKaleidoscope VITAL Steerable setting | Value Alignment Score72.11 | 42 | 3mo ago | ||
| VITAL Distributional MORALCHOICE | VISPA | JS Distance0.132 | 42 | 3mo ago | |
| GLOBALOPINIONQA VITAL Distributional 6 values | VISPA | JS Distance0.18 | 42 | 3mo ago | |
| BeaverTails (test) | Value Alignment Score59.9 | 24 | 21d ago | ||
| WildGuard (test) | Score29.69 | 24 | 21d ago | ||
| Confucianism-4 | Q+IF | Conformity Score3.842 | 22 | 7d ago | |
| HH Balance-8 | PICACO | Conformity Score4.317 | 17 | 7d ago | |
| Harmlessness 4 | PICACO | Conformity Score4.305 | 16 | 7d ago | |
| Helpfulness 4 | Q+IF+COT | Conformity Score4.364 | 16 | 7d ago | |
| 5-country prototyping panel (BRA, CHN, DEU, JPN, USA) | DISCA (Phi-3.5-mini) | Mean MIS54.5 | 13 | 22d ago | |
| Moral Integrity Corpus (MIC) 113.8k prompt-response pairs (test) | VAS-CFA | F1 (ROUGE-L)16.92 | 12 | 2mo ago | |
| Liberalism 4 | PICACO | Conformity Score3.247 | 11 | 7d ago | |
| Ethics (test) | AEM + VM | Align Score5.57 | 10 | 3mo ago | |
| MIC (test) | AEM + VM | Align Score5.48 | 10 | 3mo ago | |
| Moral Stories (test) | AEM + VM | Align Score4.85 | 10 | 3mo ago | |
| cross-demographic (test) | Qwen3-8B-DVMap | Accuracy48.6 | 9 | 19d ago |