Short-form generation on DiffTask Short-form generation
0.64PRRMSP
Evaluation Results
| Method | Links | |
|---|---|---|
| MSPModel=Gemma-2-9B2026.04 | 0.64 | |
| HBOModel=Gemma-2-9B2026.04 | 0.64 | |
| MSPModel=Llama 3.1-8B2026.04 | 0.57 | |
| HBOModel=Llama 3.1-8B2026.04 | 0.57 | |
| HUQ-SATRMDModel=Gemma-2-9B2026.04 | 0.38 | |
| SAPLMA (mid)Model=Llama 3.1-8B2026.04 | 0.35 | |
| SAPLMA (mid)Model=Gemma-2-9B2026.04 | 0.3 | |
| HUQ-SATRMDModel=Llama 3.1-8B2026.04 | 0.26 | |
| HUQ-SATMDModel=Llama 3.1-8B2026.04 | 0.24 | |
| SATRMD-MSPModel=Gemma-2-9B2026.04 | 0.24 | |
| HUQ-SATMDModel=Gemma-2-9B2026.04 | 0.23 | |
| SATRMD-MSPModel=Llama 3.1-8B2026.04 | 0.2 | |
| SATMD-MSPModel=Llama 3.1-8B2026.04 | 0.15 | |
| SATMD-MSPModel=Gemma-2-9B2026.04 | -0.05 |