Autonomous LLM Fine-tuning on LawBench
51.2Hybrid ScoreQwen3-235B-2507
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-235B-2507Model Category=Ref Model2026.04 | 51.2 | |
| TREXResearcher Backend=Qwen3-Next-80B-Thinking, Base Model Scale=Qwen3-1.7B2026.04 | 42.1 | |
| TREXResearcher Backend=Gemini 3 Pro, Base Model Scale=Qwen3-1.7B2026.04 | 40.9 | |
| Qwen3-1.7BModel Category=Base Model2026.04 | 24.2 |