Vibe-Bench - BM25 Retrieval - Response ROUGE-L: leaderboard
Vibe-Bench tests personalization from user histories: selecting an emotion-regulation strategy, generating a personalized response, and predicting occupational preferences. The paper separates no-history, full-history, profile, retrieval and personalized fine-tuning protocols. P-SFT checkpoints retain distinct identities from their untrained backbones. This series reports Vibe-Bench - BM25 Retrieval - Response ROUGE-L from 2609.00921v1, A11.T10: Score (0-100). Higher is better.
Metric: Score (0-100). Source: arxiv.org. Saturation forecast: Estimated already saturated. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 235B A22B 2507 Instruct | 18.5 | #619 |
| 2 | Gemma 3 27B (IT) | 17.85 | #1306 |
| 3 | Mistral Small 3.2 | 17.77 | #1592 |
| 4 | Llama 3.1 70B Instruct | 17.15 | #1541 |
| 5 | Qwen 3 14B | 16.96 | #1235 |
| 6 | Llama 3.1 8B Instruct | 16.84 | #4277 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=vibe-bench-bm25-retrieval-response-rouge-l · How It Works · Data refreshed daily, snapshot 2026-11-12.