MemSyco-Bench - Personalized Memory Use (Full Dialog): leaderboard
MemSyco-Bench (Xiamen University, 2026) tests memory-induced sycophancy, where remembered user beliefs or preferences improperly steer an answer. This task asks for recommendations, advice or subjective choices that should use the valid remembered preference. The model receives the whole earlier dialogue in context with no memory system; generation accuracy in percent, higher is better.
Metric: Generation accuracy (%; correct under the reference answer and task-specific rubric; recommendation, advice and subjective-choice requests that should use the valid memory; the whole preceding multi-turn dialogue is given in context with no memory system). Source: arxiv.org. Saturation forecast: Around April 2027. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V4 Flash | 60.34 | #258 |
| 2 | Qwen 3 8B | 45.67 | #1830 |
| 3 | Llama 3.1 8B Instruct | 44 | #4246 |
| 4 | GPT-4o Mini | 39.33 | #1623 |
| 5 | Llama 3.3 70B Instruct | 36 | #1357 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=memsyco-bench-personalized-memory-use-full-dialog · How It Works · Data refreshed daily, snapshot 2026-11-13.