Text Game on Frozen Lake (test)
38.3AccuracyOPCD
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| OPCDModel=Qwen3-1.7B2026.02 | 38.3 | 66.7 | |
| Context Distill.Model=Qwen3-1.7B2026.02 | 35.2 | 65.4 | |
| In-ContextModel=Qwen3-1.7B2026.02 | 31.4 | — | |
| Base ModelModel=Qwen3-1.7B2026.02 | 6.3 | 67.3 |