Evidence Retrieval on CORTEXBENCH Bridge
38.4Hit@5Gemini 3.1 Pro
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Gemini 3.1 Pro2026.06 | 38.4 | 14.1 | |
| GPT-4o2026.06 | 33.4 | 27.9 | |
| GPT-5.42026.06 | 24.4 | 28.6 | |
| DeepSeek-V4-Flash2026.06 | 23.1 | 31.6 | |
| DeepSeek-V3.22026.06 | 22.7 | 31.4 | |
| Gemini 3 Pro2026.06 | 22.5 | 55 | |
| Claude Opus 4.62026.06 | 20.6 | 40.8 | |
| Claude Sonnet 4.62026.06 | 20.3 | 41.1 |