Vibe-Bench - BM25 Retrieval - Strategy Weighted-F1: 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 - Strategy Weighted-F1 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 | Mistral Small 3.2 | 22.55 | #1592 |
| 2 | Qwen 3 235B A22B 2507 Instruct | 18.45 | #619 |
| 3 | Qwen 3 14B | 17.53 | #1235 |
| 4 | Llama 3.1 70B Instruct | 16.82 | #1541 |
| 5 | Llama 3.1 8B Instruct | 16.64 | #4277 |
| 6 | Gemma 3 27B (IT) | 14.68 | #1306 |
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-strategy-weighted-f1 · How It Works · Data refreshed daily, snapshot 2026-11-12.