Database on LiC
95.5Concat ScoreGemini-2.5-Flash
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Gemini-2.5-Flash2025.10 | 95.5 | 51.3 | 53.7 | 32.8 | |
| Gemini-2.5-Pro2025.10 | 94.9 | 43.8 | 46.2 | 33.1 | |
| GPT-4o2025.10 | 91.7 | 42.3 | 46.1 | 29.4 | |
| GPT4o-mini2025.10 | 90.7 | 40.2 | 44.3 | — | |
| DeepSeek-R12025.10 | 89.9 | 31.5 | 35 | — | |
| Llama3.3-70B2025.10 | 87.9 | 35.4 | 40.3 | — | |
| GPT-4.12025.10 | 86.5 | 46 | 53.2 | 35.8 | |
| GPT-o32025.10 | 83.3 | 35.4 | 42.5 | — | |
| Llama4-Scout2025.10 | 81.5 | 27.1 | 33.3 | — | |
| Claude3.7-Sonnet2025.10 | 81.5 | 34.9 | 42.8 | — | |
| Phi-42025.10 | 79.6 | 33.1 | 41.6 | — | |
| Qwen3-8BOptimization=RLAAR2025.10 | 77.6 | 63.2 | 81.4 | 69.7 | |
| Claude3-Haiku2025.10 | 76.5 | 31.8 | 41.6 | — | |
| Qwen3-8Bsource=Grattafiori et al.2025.10 | 76 | 46.1 | 60.6 | 31.1 | |
| Qwen2.5-7BOptimization=RLAAR2025.10 | 74.9 | 65.4 | 87.3 | 61.3 | |
| Qwen2.5-7B2025.10 | 73.2 | 38.5 | 52.5 | 27.7 | |
| Qwen2.5-7BFinetuning=SFT, Distillation=Gemini-2.5-Pro2025.10 | 69.6 | 36.4 | 52.3 | 42.9 | |
| Qwen3-8BFinetuning=SFT, Distillation=Gemini-2.5-Pro2025.10 | 67.6 | 40.2 | 59.5 | 47.8 | |
| Qwen3-1.7Bsource=Grattafiori et al.2025.10 | 51.1 | 38.8 | 76 | 20.4 | |
| Qwen3-1.7BOptimization=RLAAR2025.10 | 47.7 | 41.1 | 92.5 | 62 | |
| OLMo22025.10 | 40.5 | 22.4 | 55.3 | — |