MemSyco-Bench - Valid Memory Selection (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 questions after the user updated, reversed or replaced an earlier preference, so only the current one should be followed. 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; requests after the user updated, reversed or replaced an earlier preference; the whole preceding multi-turn dialogue is given in context with no memory system). Source: arxiv.org. Saturation forecast: Around February 2027. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V4 Flash | 77.67 | #258 |
| 2 | Llama 3.3 70B Instruct | 35.71 | #1357 |
| 3 | Llama 3.1 8B Instruct | 30.29 | #4246 |
| 4 | Qwen 3 8B | 27.79 | #1830 |
| 5 | GPT-4o Mini | 24.29 | #1623 |
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-valid-memory-selection-full-dialog · How It Works · Data refreshed daily, snapshot 2026-11-13.