Physical Reasoning on InterPhyre Catapult (evaluation)
54AccuracyHEXA
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| HEXABackbone=GPT-OSS-120B, Search Configuration=Off2On Evolving, Seeds=502026.06 | 54 | 16.2 | 54 | |
| REACT baselineBackbone=GPT-OSS-120B, Search Configuration=Off2On Evolving, Seeds=502026.06 | 0 | 25 | — |