TriBench-Ko - Hallucination: leaderboard
The hallucination risk dimension of TriBench-Ko (2026), a Korean benchmark of attorney-reviewed yes/no items built from court decisions, where false statements carry fabricated case citations, statutory provisions or facts. 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 Hallucination risk items (fabricated case citations, statutory provisions or facts): 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: Estimated already saturated. 13 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 93.3 | #138 |
| 2 | Qwen 3.5 9B | 88.9 | #838 |
| 3 | GPT-5.4 Mini | 88.3 | #409 |
| 4 | GPT-4o | 84.3 | #754 |
| 5 | Midm-2.0-Base-Instruct | 83.3 | #1998 |
| 6 | Phi-4 | 82.6 | #1951 |
| 7 | Ministral-3-8B-Instruct-2512 | 77.2 | #1923 |
| 8 | Qwen 3 8B | 70.6 | #1890 |
| 9 | Gemma 3 12B (IT) | 68.2 | #1907 |
| 10 | EXAONE 3.5 7.8B Instruct | 64.8 | #3453 |
| 11 | kanana-1.5-8B-instruct-2505 | 56.6 | #3297 |
| 12 | Llama 3.1 8B Instruct | 37.8 | #4655 |
| 13 | A.X-3.1-Light | 33.1 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=tribench-ko-hallucination · How It Works · Data refreshed daily, snapshot 2026-11-22.