BehaviorBench (Behavioral Traces) - Belief Choice (RetrievalGen): leaderboard

Metric: Choice accuracy (%): predicted final revealed side (YES or NO) of the wallet in the target market; history interface RetrievalGen (behaviorally similar examples retrieved from a disjoint support pool of wallets); 2,000 pseudonymous prediction-market wallets reconstructed from public Polymarket metadata and Polygon on-chain logs, 22,204 chronologically held-out wallet-market test instances; chance 50, context-only baseline 53.13; higher is better. Source: arxiv.org. Saturation forecast: Around November 2027. 11 models tracked.

Top models

#ModelScore
1GPT-5.471.03
2Claude Opus 470.22
3Gemma 3 12B (IT)70.06
4Gemma 3 27B (IT)69.89
5Claude 3.7 Sonnet69.61
6GPT-5.4 Mini66.96
7GPT-5.4 Nano66.8
8Llama 4 Scout Instruct65.12
9GPT-OSS-20B62.76
10NVIDIA-Nemotron-3-Nano-30B-A3B-BF1660.79
11Ministral-3-14B-Instruct-251260.43

Interactive version: theaggregate.ai/benchmark?slug=behaviorbench-behavioral-traces-belief-choice-retrievalgen · How It Works · Data refreshed daily, snapshot 2026-09-29.