PAS Police Readiness: leaderboard
Police decision-making (PAS, Police Action Scenarios; Lee, Kim and Kim, Korea University and Korean National Police Agency, 2026): 75 operational-readiness questions an officer would rehearse before a shift, built with police experts from 1,602 Korean National Police Agency manuals. Answers are rated 1-5 for overall quality by three LLM-judge passes and two police experts against an expert reference answer; higher is better.
Metric: Overall quality score (1-5 Likert; mean of three LLM-judge and two police-expert ratings against an expert reference answer, 75 questions). Source: arxiv.org. Saturation forecast: Around February 2027. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4 | 2.69 | #996 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=pas-police-readiness · How It Works · Data refreshed daily, snapshot 2026-11-12.