Open-ended Health Reasoning on HealthBench
41.45ScoreARES-RL
Evaluation Results
| Method | Links | |
|---|---|---|
| ARES-RL2026.05 | 41.45 | |
| Webscale2026.05 | 36.08 | |
| ARES-SFT2026.05 | 35.78 | |
| CPT2026.05 | 35.04 | |
| Natural Reasoning2026.05 | 32.94 |
Loading the SOTA2 catalog…
| Method | Links | |
|---|---|---|
| ARES-RL2026.05 | 41.45 | |
| Webscale2026.05 | 36.08 | |
| ARES-SFT2026.05 | 35.78 | |
| CPT2026.05 | 35.04 | |
| Natural Reasoning2026.05 | 32.94 |