PreAct-Bench - Politics: leaderboard
Metric: Prefix Foresight F1 with the full prefix (PFF(100), binary F1 of the unethical class, 0-1 scaled to %): given an agent trajectory truncated just before its first overt unethical step, the monitor predicts whether the trajectory will turn unethical; paired ethical and unethical trajectories per task; default generation parameters; higher is better; the politics domain. Source: arxiv.org. Saturation forecast: Around April 2028. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemma 3 27B | 82.1 |
| 2 | Qwen 3 30B A3B 2507 Instruct | 81.3 |
| 3 | GPT-5.4 | 77.7 |
| 4 | Qwen 3 8B (Non-reasoning) | 74.5 |
| 5 | Qwen 3 8B (Thinking) | 74.3 |
| 6 | Gemma 3 4B | 66.7 |
| 7 | Qwen 3 0.6B (Non-reasoning) | 66.7 |
| 8 | GPT-4o | 62.1 |
| 9 | Qwen 3 30B A3B 2507 (Thinking) | 60.9 |
| 10 | Qwen 3 0.6B (Thinking) | 58.8 |
| 11 | Qwen 3 4B 2507 (Thinking) | 55.4 |
| 12 | Qwen 3 4B 2507 Instruct | 54 |
| 13 | GPT-OSS-20B | 53.5 |
| 14 | GPT-OSS-120B | 51.7 |
Interactive version: theaggregate.ai/benchmark?slug=preact-bench-politics · How It Works · Data refreshed daily, snapshot 2026-09-29.