Social reasoning on Social regime (n = 300)
51.3Exact Match AccuracySituational leadership
Evaluation Results
| Method | Links | |
|---|---|---|
| Situational leadershipModel=llama-4-scout2026.06 | 51.3 | |
| Transformational leadershipModel=llama-4-scout2026.06 | 48.3 | |
| Theory-free controlModel=llama-4-scout2026.06 | 46.7 | |
| directives-only ablationModel=llama-4-scout2026.06 | 45.3 | |
| Situational leadershipModel=gpt-oss-120b2026.06 | 45 | |
| FlatModel=llama-4-scout2026.06 | 43.3 | |
| Transformational leadershipModel=gpt-oss-120b2026.06 | 43 | |
| Transactional leadershipModel=llama-4-scout2026.06 | 43 | |
| accept-only ablationModel=llama-4-scout2026.06 | 42.7 | |
| Theory-free controlModel=gpt-oss-120b2026.06 | 42.3 | |
| Transactional leadershipModel=gpt-oss-120b2026.06 | 42.3 | |
| accept-only ablationModel=gpt-oss-120b2026.06 | 42.3 | |
| directives-only ablationModel=gpt-oss-120b2026.06 | 42 | |
| FlatModel=gpt-oss-120b2026.06 | 41.7 | |
| Transformational leadershipModel=gemma-4-31B-it2026.06 | 39.7 | |
| Theory-free controlModel=gemma-4-31B-it2026.06 | 39 | |
| Transactional leadershipModel=gemma-4-31B-it2026.06 | 38.7 | |
| accept-only ablationModel=gemma-4-31B-it2026.06 | 38.7 | |
| FlatModel=gemma-4-31B-it2026.06 | 38.3 | |
| Situational leadershipModel=gemma-4-31B-it2026.06 | 38 | |
| directives-only ablationModel=gemma-4-31B-it2026.06 | 37.3 |