Reasoning on Flawed Fictions (test)
88.71AccuracyRL-Trained
Evaluation results
| Method | Links | ||
|---|---|---|---|
| RL-TrainedTraining Status=Trained2025.12 | 88.71 | 114.53 | |
| LiteReasonTraining Status=Trained2025.12 | 87.42 | 34.01 | |
| MoITraining Status=Untrained2025.12 | 58.06 | 401.73 | |
| Soft ThinkingTraining Status=Untrained2025.12 | 57.42 | 360.9 | |
| Qwen2.5-7BTraining Status=Untrained2025.12 | 57.26 | 400.04 | |
| CoLaRTraining Status=Trained2025.12 | 53.71 | 226.14 | |
| COCONUTTraining Status=Trained2025.12 | 50.65 | 268.07 |