Agentic Search on BrowseComp-Plus
66.5AccuracySolution Aggregation
Evaluation Results
| Method | Links | |
|---|---|---|
| Solution AggregationModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 66.5 | |
| FINEVERIFYModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 66.5 | |
| Weighted VotingModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 64 | |
| Best-of-NModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 63.5 | |
| Confidence VerifyModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 63 | |
| Majority VotingModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 62.5 | |
| FINEVERIFYModel=GPT-5-mini, Sampling Trajectories=42026.05 | 60.5 | |
| Pass@1Model=Gemini-3-flash, Sampling Trajectories=12026.05 | 60.5 | |
| Best-of-NModel=GPT-5-mini, Sampling Trajectories=42026.05 | 59.5 | |
| Solution AggregationModel=GPT-5-mini, Sampling Trajectories=42026.05 | 59.5 | |
| Confidence VerifyModel=GPT-5-mini, Sampling Trajectories=42026.05 | 58 | |
| Weighted VotingModel=GPT-5-mini, Sampling Trajectories=42026.05 | 57.5 | |
| Pass@1Model=GPT-5-mini, Sampling Trajectories=12026.05 | 49.5 | |
| Majority VotingModel=GPT-5-mini, Sampling Trajectories=42026.05 | 49.5 |