Autonomous Research Paper Generation on LIAR
71ChatGPT ScoreClaw AI Lab
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Claw AI LabMain Model=GPT-5.4, Coding Model=GPT-5.4, Figure-generation Model=Gemini-3-Pro-Image-Preview, Fallback Models=Qwen3.5-Plus/Qwen-Plus2026.05 | 71 | 73 | 15.5 | |
| AutoResearchClawMain Model=GPT-5.4, Image Model=Gemini-2.5-Pro-Flash-Image, Fallback Models=GPT-4o/GPT-4o-mini2026.05 | 49 | 64 | — |