Function-name selection on Seal-Tools single-call K=5 (294 tasks/model)
93.2AccuracyLlama-3.1-8B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Llama-3.1-8Bsetting=Gold-free S22026.06 | 93.2 | 16.7 | |
| Qwen2.5-1.5Bsetting=Gold-free S22026.06 | 91.5 | 11.9 | |
| Pooledsetting=Gold-free S22026.06 | 90.7 | 14.9 | |
| Qwen2.5-7Bsetting=Gold-free S22026.06 | 87.4 | 16 | |
| Qwen2.5-1.5Bsetting=Base2026.06 | 79.6 | — | |
| Llama-3.1-8Bsetting=Base2026.06 | 76.5 | — | |
| Pooledsetting=Base2026.06 | 75.9 | — | |
| Qwen2.5-7Bsetting=Base2026.06 | 71.4 | — |