Personalization on GOQA
85.2AccuracyRPM
Evaluation Results
| Method | Links | |
|---|---|---|
| RPMBackbone=GPT-4o-mini, +CoT=false, Reasoning generation=enabled2025.05 | 85.2 | |
| PAGBackbone=GPT-4o-mini, +CoT=true2025.05 | 82 | |
| RPMBackbone=GPT-4o-mini, +CoT=false, Reasoning generation=disabled2025.05 | 82 | |
| HYDRABackbone=GPT-4o-mini, +CoT=true2025.05 | 80.6 | |
| RAGBackbone=GPT-4o-mini, +CoT=true2025.05 | 80 | |
| HYDRABackbone=GPT-4o-mini, +CoT=false2025.05 | 80 | |
| FermiBackbone=GPT-4o-mini, +CoT=false2025.05 | 80 | |
| PAGBackbone=GPT-4o-mini, +CoT=false2025.05 | 79.5 | |
| RAGBackbone=GPT-4o-mini, +CoT=false2025.05 | 77.3 | |
| ICLBackbone=GPT-4o-mini, +CoT=false2025.05 | 69.5 | |
| ICLBackbone=GPT-4o-mini, +CoT=true2025.05 | 68.1 | |
| FermiBackbone=GPT-4o-mini, +CoT=true2025.05 | 65.9 | |
| Zero-shotBackbone=GPT-4o-mini, +CoT=false2025.05 | 56.2 | |
| Zero-shotBackbone=GPT-4o-mini, +CoT=true2025.05 | 55.7 |