Climate Discovery on ClimaBench 2000–2024
7.38AE (Absolute Error)DeepSeek-R1
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| DeepSeek-R1Backbone=DeepSeek-R1-671B2026.04 | 7.38 | 4.88 | 8.75 | 4.62 | 6.41 | |
| Qwen3-235BBackbone=Qwen3-235B2026.04 | 7.42 | 4.92 | 8.72 | 4.68 | 6.45 | |
| GPT-4oBackbone=GPT-4o2026.04 | 7.55 | 3.92 | 8.42 | 5.25 | 6.29 | |
| ResearchAgentBackbone=GPT-4o2026.04 | 8.05 | 6.85 | 8.85 | 7.42 | 7.79 | |
| DS-AgentBackbone=GPT-4o2026.04 | 8.24 | 7.12 | 8.75 | 7.51 | 7.91 | |
| ResearchAgentBackbone=DeepSeek-R1-671B2026.04 | 8.24 | 7.21 | 8.85 | 7.12 | 7.86 | |
| ResearchAgentBackbone=Qwen3-235B2026.04 | 8.28 | 7.25 | 8.88 | 7.15 | 7.89 | |
| DS-AgentBackbone=DeepSeek-R1-671B2026.04 | 8.35 | 7.02 | 8.82 | 7.35 | 7.89 | |
| DS-AgentBackbone=Qwen3-235B2026.04 | 8.38 | 7.05 | 8.85 | 7.38 | 7.92 | |
| Agent LaboratoryBackbone=GPT-4o2026.04 | 8.62 | 6.42 | 8.68 | 5.68 | 7.35 | |
| DeepAnalyzeBackbone=GPT-4o2026.04 | 8.62 | 6.28 | 8.75 | 6.22 | 7.34 | |
| DeepAnalyzeBackbone=DeepSeek-R1-671B2026.04 | 8.62 | 6.28 | 8.75 | 6.22 | 7.34 | |
| DeepAnalyzeBackbone=Qwen3-235B2026.04 | 8.62 | 6.28 | 8.75 | 6.22 | 7.34 | |
| Agent LaboratoryBackbone=DeepSeek-R1-671B2026.04 | 8.75 | 6.12 | 8.8 | 6.12 | 7.45 | |
| Agent LaboratoryBackbone=Qwen3-235B2026.04 | 8.78 | 6.15 | 8.82 | 6.15 | 7.48 | |
| ClimAgentBackbone=GPT-4o2026.04 | 9.22 | 7.45 | 9.12 | 8.58 | 8.92 | |
| ClimAgentBackbone=DeepSeek-R1-671B2026.04 | 9.62 | 8.35 | 9.18 | 8.65 | 8.95 | |
| ClimAgentBackbone=Qwen3-235B2026.04 | 9.65 | 8.38 | 9.22 | 8.68 | 8.98 |