BehaviorBench (Behavioral Traces) - Trade Amount Error (DirectGen): leaderboard
BehaviorBench (Yang et al., 2026) tests whether a model can predict what a real pseudonymous user will do, using 2,000 wallets reconstructed from public prediction-market and on-chain records with chronologically held-out targets. The model sees the recent raw history of the wallet (DirectGen). The board is the median absolute error of the predicted transaction amount, in paper units and unbounded; lower is better.
Metric: Median absolute error of the predicted trade amount (paper units, unbounded); history interface DirectGen (the target wallet's most recent visible history, up to 30 events for Belief); 2,000 pseudonymous prediction-market wallets reconstructed from public Polymarket metadata and Polygon on-chain logs, about 59,560 chronologically held-out test transactions; unparsable outputs count as failures; context-only baseline 31.79; lower is better. Source: arxiv.org. Saturation forecast: Not forecast. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 3 12B (IT) | 8.34 | #1837 |
| 2 | Gemma 3 27B (IT) | 9 | #1289 |
| 3 | Llama 4 Scout Instruct | 9.14 | #1543 |
| 4 | Claude 3.7 Sonnet | 9.49 | #458 |
| 5 | GPT-5.4 | 9.63 | #123 |
| 6 | GPT-5.4 Mini | 9.8 | #389 |
| 7 | NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 | 9.95 | #1767 |
| 8 | GPT-OSS-20B | 10.4 | #1147 |
| 9 | GPT-5.4 Nano | 11.37 | #635 |
| 10 | Claude Opus 4 | 12.47 | #322 |
| 11 | Ministral-3-14B-Instruct-2512 | 21.99 | #1593 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=behaviorbench-behavioral-traces-trade-amount-error-directgen · How It Works · Data refreshed daily, snapshot 2026-11-13.