FinED-Bench - General Knowledge Errors: leaderboard
Metric: F1 (%; general knowledge errors: illegal time, redundant statements, value format errors, missing numerical and non-numerical attribute values; a predicted error counts as correct when the extracted sentence matches or contains the annotated sentence and the error type is classified correctly; the paper's error-detection prompt over each document, reasoning models with thinking on unless marked no thinking). Source: arxiv.org. Saturation forecast: Estimated already saturated. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o | 52.33 |
| 2 | Qwen 3 14B | 51.41 |
| 3 | Qwen 3 8B | 47.6 |
| 4 | Qwen 3 14B (Non-reasoning) | 33.72 |
| 5 | GPT-4o Mini | 29.69 |
| 6 | DeepSeek R1 0528 Qwen3 8B | 29.09 |
| 7 | Qwen 3 8B (Non-reasoning) | 24.93 |
| 8 | Qwen 2.5 7B Instruct | 14.67 |
| 9 | GPT-3.5 Turbo | 14.05 |
Interactive version: theaggregate.ai/benchmark?slug=fined-bench-general-knowledge-errors · How It Works · Data refreshed daily, snapshot 2026-09-26.