LexKairos - Event Relation Discerning: leaderboard
The Event Relation Discerning sub-task of LexKairos (Tsinghua University, 2026): classify the temporal relation between two events of a judicial case. F1 score in percent over 450 items in the vanilla setting (direct answer); higher is better.
Metric: F1 score (%; 450 items: classify the temporal relation between two events of a judicial case; vanilla setting, the model answers directly without extra reasoning steps (thinking disabled where the model allows it); temperature 0, 8,000 output tokens). Source: arxiv.org. Saturation forecast: Around February 2027. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 (Non-reasoning) | 90.58 | #123 (GPT-5.4) |
| 2 | DeepSeek R1 | 87.37 | #473 |
| 3 | Gemini 3 Flash (Minimal) | 82.53 | #164 (Gemini 3 Flash) |
| 4 | Qwen 3.5 9B (Non-reasoning) | 69.94 | #803 (Qwen 3.5 9B) |
| 5 | DeepSeek V4 Flash (Non-reasoning) | 54.13 | #258 (DeepSeek V4 Flash) |
| 6 | Llama 3.1 8B | 28.04 | #4861 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=lexkairos-event-relation-discerning · How It Works · Data refreshed daily, snapshot 2026-11-13.