Mathematical Claim Generation on Mathematical Claim Generation LLM Judge Evaluation pre-2024 GPT
3.45Content ScoreCOMPOSE
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| COMPOSEBackbone=DeepSeek2026.05 | 3.45 | 3.52 | 3.1 | 3.23 | 3.52 | 3.36 | |
| Prompt-only2026.05 | 2.88 | 3.16 | 3.21 | 3.02 | 3.21 | 3.1 | |
| GOAI-GPT2026.05 | 2.52 | 2.69 | 3.24 | 2.9 | 3.29 | 2.93 | |
| GIANTS2026.05 | 1.91 | 2.75 | 3.27 | 2.53 | 3.28 | 2.75 | |
| Text-onlyTraining=LoRA2026.05 | 1.86 | 1.91 | 2.42 | 2.04 | 2.43 | 2.13 | |
| ResearchAgent-GPT2026.05 | 1.74 | 2.1 | 2.86 | 2.21 | 2.99 | 2.38 | |
| Paper-graph-only2026.05 | 1.59 | 1.57 | 1.71 | 1.46 | 1.71 | 1.61 | |
| FutureGen2026.05 | 1.4 | 1.79 | 2.75 | 1.85 | 2.77 | 2.11 |