TriBench-Ko - Adjudicative Overreach: leaderboard
The adjudicative overreach risk dimension of TriBench-Ko (2026), a Korean benchmark of attorney-reviewed yes/no items built from court decisions, where false statements carry statements that presuppose judicial authority reserved to the judge. The score is macro-F1 on a 0-100 scale over the atomic binary judgments, zero-shot at temperature 0; higher is better.
Metric: Macro-F1 (0-100) on the TriBench-Ko Adjudicative Overreach risk items (statements that presuppose judicial authority reserved to the judge): yes/no verification of statements about Korean court decisions, computed over the atomic binary judgments, zero-shot, temperature 0, at most 64 output tokens; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 13 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 9B | 88.2 | #828 |
| 2 | GPT-5.4 | 87.7 | #135 |
| 3 | Midm-2.0-Base-Instruct | 84.4 | #1988 |
| 4 | GPT-5.4 Mini | 82.9 | #406 |
| 5 | Ministral-3-8B-Instruct-2512 | 79.4 | #1907 |
| 6 | Phi-4 | 74.6 | #1931 |
| 7 | Gemma 3 12B (IT) | 71.6 | #1887 |
| 8 | GPT-4o | 66.4 | #751 |
| 9 | kanana-1.5-8B-instruct-2505 | 61.2 | #3315 |
| 10 | Qwen 3 8B | 50.5 | #1873 |
| 11 | EXAONE 3.5 7.8B Instruct | 46.2 | #3410 |
| 12 | Llama 3.1 8B Instruct | 41.2 | #4600 |
| 13 | A.X-3.1-Light | 37.5 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=tribench-ko-adjudicative-overreach · How It Works · Data refreshed daily, snapshot 2026-11-21.