ConvRe - Text2Re Hard: leaderboard
The "Text2Re Hard" subtask of ConvRe. ConvRe evaluates language models' understanding of converse relations through Re2Text and Text2Re tasks in easy and hard settings.
Metric: Accuracy (%). Source: huggingface.co. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | text-davinci-003 | 34.8 | #1146 |
| 2 | GPT-4 (0314) | 17.1 | #574 |
Interactive version: theaggregate.ai/benchmark?slug=convre-text2re-hard · How It Works · Data refreshed daily, snapshot 2026-10-20.