Reasoning on FOL→GSM
95.1AccuracyDIN-ICL
Evaluation Results
| Method | Links | |
|---|---|---|
| DIN-ICLModel Series=Gemma-3, Parameters=27B2026.04 | 95.1 | |
| DIN-ICLModel Series=Qwen-3, Parameters=32B2026.04 | 95 | |
| ZERO-SHOTModel Series=Qwen-3, Parameters=32B2026.04 | 94.6 | |
| X-ICLModel Series=Qwen-3, Parameters=32B2026.04 | 94.6 | |
| DIN-ICLModel Series=Qwen-3, Parameters=8B2026.04 | 94.6 | |
| ZERO-SHOTModel Series=Gemma-3, Parameters=27B2026.04 | 94.3 | |
| ZERO-SHOTModel Series=Gemma-3, Parameters=12B2026.04 | 93.7 | |
| DIN-ICLModel Series=Gemma-3, Parameters=12B2026.04 | 93.7 | |
| ZERO-SHOTModel Series=Qwen-2.5, Parameters=14B2026.04 | 93.4 | |
| DIN-ICLModel Series=Qwen-2.5, Parameters=14B2026.04 | 93.4 | |
| ZERO-SHOTModel Series=Qwen-3, Parameters=8B2026.04 | 93.3 | |
| X-ICLModel Series=Qwen-3, Parameters=8B2026.04 | 93.1 | |
| X-ICLModel Series=Gemma-3, Parameters=27B2026.04 | 93.1 | |
| X-ICLModel Series=Qwen-2.5, Parameters=14B2026.04 | 92.9 | |
| X-ICLModel Series=Gemma-3, Parameters=12B2026.04 | 92.6 | |
| DIN-ICLModel Series=Qwen-2.5, Parameters=32B2026.04 | 92.1 | |
| ZERO-SHOTModel Series=Qwen-2.5, Parameters=32B2026.04 | 91.7 | |
| X-ICLModel Series=Qwen-2.5, Parameters=32B2026.04 | 91.6 | |
| ZERO-SHOTModel Series=Qwen-2.5, Parameters=7B2026.04 | 91.1 | |
| DIN-ICLModel Series=Qwen-3, Parameters=14B2026.04 | 91 | |
| ZERO-SHOTModel Series=Qwen-3, Parameters=14B2026.04 | 89.9 | |
| X-ICLModel Series=Qwen-3, Parameters=14B2026.04 | 89.9 | |
| DIN-ICLModel Series=Qwen-2.5, Parameters=7B2026.04 | 89.7 | |
| X-ICLModel Series=Qwen-2.5, Parameters=7B2026.04 | 89.6 | |
| ZERO-SHOTModel Series=LLaMA-3.1, Parameters=8B2026.04 | 81.6 | |
| DIN-ICLModel Series=LLaMA-3.1, Parameters=8B2026.04 | 81.5 | |
| X-ICLModel Series=LLaMA-3.1, Parameters=8B2026.04 | 81.4 |