M3R-Bench - Source Prediction: leaderboard
Metric: BERTScore F1 (0-100) of the predicted central source concept against the human annotation, predicted jointly with metaphor presence, target and sentiment. Source: arxiv.org. Saturation forecast: Around 2029. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 71.35 |
| 2 | GPT-4o | 71.23 |
| 3 | GPT-5.5 | 67.02 |
| 4 | Qwen 3 VL 32B Instruct | 66.53 |
| 5 | Qwen 3 VL 8B Instruct | 66.4 |
| 6 | Kimi K2.5 | 63.51 |
Interactive version: theaggregate.ai/benchmark?slug=m3r-bench-source-prediction · How It Works · Data refreshed daily, snapshot 2026-09-26.