Science Question Answering on ARC (Reusability Score Focus)
49.45Reusability ScorePhi
Evaluation Results
| Method | Links | |
|---|---|---|
| PhiExecutor=Strong Comm.2026.02 | 49.45 | |
| LlamaExecutor=Strong Comm.2026.02 | 44.64 | |
| PhiExecutor=Full Comm.2026.02 | 43.4 | |
| LlamaExecutor=Full Comm.2026.02 | 42.72 | |
| LlamaExecutor=Weak Comm.2026.02 | 40.79 | |
| PhiExecutor=Weak Comm.2026.02 | 37.34 | |
| GemmaExecutor=Strong Comm.2026.02 | 36.95 | |
| GemmaExecutor=Full Comm.2026.02 | 35.88 | |
| GemmaExecutor=Weak Comm.2026.02 | 34.81 | |
| R1Executor=Strong Comm.2026.02 | 33.04 | |
| R1Executor=Full Comm.2026.02 | 32.3 | |
| R1Executor=Weak Comm.2026.02 | 31.55 |