Context Learning Task-Solving on CL-Bench
15.8Overall ScoreContextGuard
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| ContextGuardModel=Qwen3.5-9B2026.05 | 15.8 | 17.65 | 14.13 | 16.14 | 13.57 | |
| ContextGuardModel=Qwen3.5-4B2026.05 | 13.85 | 14.48 | 13.07 | 14.01 | 13.57 | |
| Self-RefineModel=Qwen3.5-4B2026.05 | 10.48 | 11.76 | 10.42 | 9.34 | 9.05 | |
| BaselineModel=Qwen3.5-9B2026.05 | 10.43 | 11.31 | 10.25 | 10.62 | 7.54 | |
| BaselineModel=Qwen3.5-4B2026.05 | 9.64 | 9.8 | 9.9 | 9.13 | 9.55 |