KoNeoBench - Definition with Examples: leaderboard
KoNeoBench evaluates Korean neologism understanding: contextual identification, component source-word recovery, semantic and specialist-domain categorization, and definition generation. Table 6 identifies the instruction-tuned checkpoints behind shortened result labels. This series reports KoNeoBench - Definition with Examples from 2609.19916v1, S4.T1: Definition Judge Score (0-10). Higher is better.
Metric: Definition Judge Score (0-10). Source: arxiv.org. Saturation forecast: Around February 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 8.69 | #121 |
| 2 | GPT-4.1 | 8.47 | #492 |
| 3 | Solar Pro 3 | 7.79 | #1694 |
| 4 | EXAONE 3.5 7.8B Instruct | 7.53 | #3050 |
| 5 | Qwen 3.5 9B | 7.05 | #803 |
| 6 | Qwen 2.5 7B Instruct | 6.81 | #2583 |
| 7 | Llama 3.1 8B Instruct | 6.31 | #4277 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=koneobench-definition-with-examples · How It Works · Data refreshed daily, snapshot 2026-11-12.