Success Rate on Bongard-LOGO
80Success RateTwo-stage pipeline (o1 + GPT-4o)
Evaluation Results
| Method | Links | |
|---|---|---|
| Two-stage pipeline (o1 + GPT-4o)Pipeline=Two-stage, Perception=o1, Reasoning=GPT-4o2025.12 | 80 | |
| o1 (Standard One-Stage)Pipeline=One-stage, Model=o12025.12 | 78 | |
| Two-stage pipeline (GPT-4o + GPT-4o)Pipeline=Two-stage, Perception=GPT-4o, Reasoning=GPT-4o2025.12 | 73 | |
| GPT-4o (Standard One-Stage)Pipeline=One-stage, Model=GPT-4o2025.12 | 62 |