Code on LiC Code
97.1Concat ScoreDeepSeek-R1
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| DeepSeek-R12025.10 | 97.1 | 70.9 | 73 | — | |
| Gemini-2.5-Pro2025.10 | 95.7 | 68.1 | 71.2 | 30.2 | |
| Gemini-2.5-Flash2025.10 | 92.5 | 68.3 | 73.8 | 26.8 | |
| GPT-4.12025.10 | 88.7 | 72.6 | 81.8 | 27.4 | |
| GPT-o32025.10 | 87.2 | 53 | 60.8 | — | |
| GPT-4o2025.10 | 82.9 | 61.3 | 73.9 | 29.8 | |
| Claude3.7-Sonnet2025.10 | 76.2 | 65.6 | 86.1 | — | |
| GPT4o-mini2025.10 | 66.7 | 50.3 | 75.4 | — | |
| Qwen3-8BOptimization=RLAAR2025.10 | 63.2 | 44.8 | 70.9 | 72.3 | |
| Llama4-Scout2025.10 | 60.3 | 46.4 | 76.9 | — | |
| Qwen3-8Bsource=Grattafiori et al.2025.10 | 60 | 35 | 58.3 | 28.9 | |
| Qwen2.5-7BOptimization=RLAAR2025.10 | 58.7 | 41.2 | 70.2 | 72 | |
| Qwen3-8BFinetuning=SFT, Distillation=Gemini-2.5-Pro2025.10 | 58.4 | 34.2 | 58.6 | 42.9 | |
| Qwen2.5-7BFinetuning=SFT, Distillation=Gemini-2.5-Pro2025.10 | 56.4 | 33.9 | 60.1 | 47.7 | |
| Qwen2.5-7B2025.10 | 55 | 31 | 56.4 | 26.2 | |
| Llama3.3-70B2025.10 | 52.7 | 51.6 | 97.9 | — | |
| Phi-42025.10 | 48.4 | 39.1 | 80.8 | — | |
| Qwen3-1.7BOptimization=RLAAR2025.10 | 37 | 22.9 | 61.94 | 69.3 | |
| Claude3-Haiku2025.10 | 36.3 | 31.5 | 86.8 | — | |
| Qwen3-1.7Bsource=Grattafiori et al.2025.10 | 33.4 | 15 | 44.9 | 19.8 | |
| OLMo22025.10 | 16.3 | 14.4 | 88.3 | — |