Weather Captioning on Weather Captioning Human Evaluation Set
8.52SA ScoreWeatherTGD
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| WeatherTGDBackbone=Qwen3-Next-80B2026.03 | 8.52 | 8.42 | 8.68 | 8.54 | 78 | |
| AgentVerseBackbone=Qwen3-Next-80B2026.03 | 7.32 | 6.98 | 7.18 | 7.14 | 75 | |
| LLM-DebateBackbone=Qwen3-Next-80B2026.03 | 7.24 | 7.08 | 6.98 | 7.12 | 76 | |
| MADBackbone=Qwen3-Next-80B2026.03 | 7.16 | 7.02 | 6.9 | 7.02 | 77 | |
| CAMELBackbone=Qwen3-Next-80B2026.03 | 7.1 | 6.88 | 6.76 | 6.9 | 78 | |
| AutoGenBackbone=Qwen3-Next-80B2026.03 | 6.92 | 6.66 | 6.8 | 6.78 | 79 | |
| Self-ConsistencyBackbone=Qwen3-Next-80B2026.03 | 6.88 | 6.62 | 6.94 | 6.78 | 80 | |
| VanillaBackbone=Qwen3-Next-80B2026.03 | 5.7 | 5.48 | 5.62 | 5.6 | 82 |