Agricultural Reasoning on AgroBench 1.0 (test)
73.84Choice AccuracyQwen3-32B-LoRA-Reflective
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Qwen3-32B-LoRA-ReflectiveBackbone=Qwen3-32B, Fine-tuning=LoRA, Reasoning Mode=Agro-Reflective2026.02 | 73.84 | — | — | — | — | — | |
| Qwen3-32B-LoRABackbone=Qwen3-32B, Fine-tuning=LoRA, Category=Base Models (LoRA Tuned)2026.02 | 67.25 | — | — | — | — | — | |
| Qwen3-8B-LoRABackbone=Qwen3-8B, Fine-tuning=LoRA, Category=Base Models (LoRA Tuned)2026.02 | 57.55 | — | — | — | — | — | |
| DeepSeek-V3Category=Open-Source Baselines2026.02 | 57.43 | — | — | — | — | — | |
| Gemini-2.0-FlashCategory=Proprietary Models2026.02 | 55.33 | — | — | — | — | — | |
| Qwen3-8B-LoRA-ReflectiveBackbone=Qwen3-8B, Fine-tuning=LoRA, Reasoning Mode=Agro-Reflective2026.02 | 41.34 | — | — | — | — | — | |
| Qwen3-32BBackbone=Qwen3-32B, Category=Open-Source Baselines2026.02 | 39.67 | — | — | — | — | — | |
| GPT-4oCategory=Proprietary Models2026.02 | 36.75 | — | — | — | — | — | |
| Yi-1.5-34B-ChatCategory=Open-Source Baselines2026.02 | 35.38 | — | — | — | — | — |