Reasoning on OERQA
69.83LS ScoreChatAD*
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ChatAD*Backbone=Qwen2.5-7B, Enhancement=TKTO2026.01 | 69.83 | 99.99 | |
| TimeMQA-Qwen2.5-7BBackbone=Qwen2.5-7B2026.01 | 66.03 | 99.97 | |
| TimeMQA-Mistral-7BBackbone=Mistral-7B2026.01 | 64.08 | 99.49 | |
| ChatAD-Mistral-7BBackbone=Mistral-7B2026.01 | 63.31 | 99.99 | |
| ChatAD-Llama3-8BBackbone=Llama3-8B2026.01 | 55.64 | 95.82 | |
| TimeMQA-Llama3-8BBackbone=Llama3-8B2026.01 | 55.57 | 91.75 | |
| ChatAD-Qwen2.5-7BBackbone=Qwen2.5-7B2026.01 | 55.45 | 99.98 | |
| DS-Qwen2.5-7BDistillation=DeepSeek-R1-Distill, Backbone=Qwen2.5-7B2026.01 | 49.93 | 95.81 | |
| Llama3-8B-InstructBackbone=Llama3-8B2026.01 | 49.11 | 99.87 | |
| DS-Llama3-8BDistillation=DeepSeek-R1-Distill, Backbone=Llama3-8B2026.01 | 47.18 | 97.87 | |
| Qwen2.5-7B-InstructBackbone=Qwen2.5-7B2026.01 | 46.49 | 99.96 | |
| Qwen3-8BBackbone=Qwen3-8B2026.01 | 43.32 | 99.88 | |
| Mistral-7B-InstructBackbone=Mistral-7B2026.01 | 38.92 | 95.3 |