Crop Disease Knowledge QA on CDDMBench 1.0 (test)
84.5QA ScoreGpt-5-Nano
Evaluation Results
| Method | Links | |
|---|---|---|
| Gpt-5-NanoGenerated Explanation Caption Source=GPT-5-mini, Strategy=llm-as-a-judge2025.12 | 84.5 | |
| Gpt-5-NanoGenerated Explanation Caption Source=GPT-5-mini, Strategy=Explanational Caption2025.12 | 84 | |
| Gpt-5-NanoGenerated Explanation Caption Source=Qwen2.5-VL-72B-Instruct, Strategy=llm-as-a-judge2025.12 | 76 | |
| Gpt-5-NanoGenerated Explanation Caption Source=GPT-5-mini, Strategy=few-shot2025.12 | 76 | |
| Gpt-5-NanoGenerated Explanation Caption Source=Qwen2.5-VL-72B-Instruct, Strategy=Explanational Caption2025.12 | 75.5 | |
| Gpt-5-NanoGenerated Explanation Caption Source=Qwen2.5-VL-72B-Instruct, Strategy=few-shot2025.12 | 74.5 | |
| Gpt-5-NanoGenerated Explanation Caption Source=Qwen2.5-VL-72B-Instruct, Strategy=Our Baseline2025.12 | 65 | |
| Gpt-5-NanoGenerated Explanation Caption Source=GPT-5-mini, Strategy=Our Baseline2025.12 | 65 | |
| Qwen-VL-ChatGenerated Explanation Caption Source=Qwen2.5-VL-72B-Instruct, Strategy=llm-as-a-judge2025.12 | 51 | |
| Qwen-VL-ChatGenerated Explanation Caption Source=Qwen2.5-VL-72B-Instruct, Strategy=few shot2025.12 | 50 | |
| Qwen-VL-ChatGenerated Explanation Caption Source=GPT-5-mini, Strategy=llm-as-a-judge2025.12 | 49.5 | |
| Qwen-VL-ChatGenerated Explanation Caption Source=GPT-5-mini, Strategy=few-shot2025.12 | 49 | |
| Qwen-VL-ChatGenerated Explanation Caption Source=Qwen2.5-VL-72B-Instruct, Strategy=Explanational Caption2025.12 | 46.5 | |
| Qwen-VL-ChatGenerated Explanation Caption Source=GPT-5-mini, Strategy=Explanational Caption2025.12 | 44 | |
| Qwen-VL-ChatGenerated Explanation Caption Source=Qwen2.5-VL-72B-Instruct, Strategy=Our Baseline2025.12 | 41.5 | |
| Qwen-VL-ChatGenerated Explanation Caption Source=Qwen2.5-VL-72B-Instruct, Strategy=Liu et al. (2024) Baseline2025.12 | 41 |