Inferential Calculation on TSRBench
77.14AccuracyVeriTime
Evaluation results
| Method | Links | |
|---|---|---|
| VeriTimeBase Model=Qwen3-4B-Instruct, Training Protocol=SFT+RL2026.02 | 77.14 | |
| ChatTSBase Model=Qwen3-4B-Instruct, Training Protocol=SFT2026.02 | 72.38 | |
| VeriTimeBase Model=Qwen2.5-3B-Instruct, Training Protocol=SFT+RL2026.02 | 68.57 | |
| GPT-4o-miniModel Type=General LLM2026.02 | 65.71 | |
| Qwen3-4B-InstructTraining Protocol=Base2026.02 | 62.85 | |
| Qwen2.5-7B-instructModel Type=General LLM2026.02 | 55.24 | |
| ChatTSBase Model=Qwen2.5-3B-Instruct, Training Protocol=SFT2026.02 | 54.28 | |
| DeepSeek-R1-Distill-Qwen-7BModel Type=General LLM2026.02 | 49.52 | |
| Time-R1Base Model=Qwen2.5-7B2026.02 | 40.35 | |
| Meta-Llama3-8B-InstructModel Type=General LLM2026.02 | 33.33 | |
| Time-MQABase Model=Llama3-8B2026.02 | 32.56 | |
| Qwen2.5-3B-InstructTraining Protocol=Base2026.02 | 31.43 | |
| Time-MQABase Model=Qwen2.5-7B2026.02 | 30.48 | |
| Mistral-7B-v0.3Model Type=General LLM2026.02 | 28.57 | |
| Time-MQABase Model=Mistral-7B2026.02 | 24.62 |