Reasoning quality evaluation on PROOFWRITER
0.339Somers' DMarODE
Evaluation Results
| Method | Links | |
|---|---|---|
| MarODE2026.03 | 0.339 | |
| MarODE(αβ)2026.03 | 0.3386 | |
| MarODE_QUALITY(β)2026.03 | 0.24 | |
| MarODE(βγ)2026.03 | 0.2382 | |
| MarODE(αγ)2026.03 | 0.2157 | |
| MarODE_COHERENCE(α)2026.03 | 0.2141 | |
| ROSCOE_MEAN2026.03 | 0.2115 | |
| ROSCOE-SA2026.03 | 0.1703 | |
| ROSCOE-SS2026.03 | 0.1544 | |
| MarODE_EVIDENCE(γ)2026.03 | 0.1369 | |
| ROSCOE-LC2026.03 | 0.0971 | |
| ROSCOE-LI2026.03 | 0.0539 | |
| LLM_as_a_Judge2026.03 | 0 | |
| Local_and_Global_Coherence2026.03 | -0.0067 | |
| ReCEval2026.03 | -0.0218 |