MARI:面向 LLM 推荐的蒸馏结构化决策记忆
Get the story
2026年10月9日,arXiv Computation and Language 发布论文《MARI:面向 LLM 推荐的蒸馏结构化决策记忆》(一手来源)。论文提出 MARI(Memory-Augmented Recommendation with Interpretability),核心方法是离线 Post-Hoc Decision Distillation:将异构行为数据与用户内容蒸馏为结构化决策记忆(SDM),并存入 Decision Memory Bank(DMB);在线阶段检索相关 SDM,以增强 LLM 的推理能力。目前公开信息仅涉及该论文的方法框架,尚无实验结果、代码开源或第三方复现报道。
Generated from reports · updated 3 hr ago
Timeline
Follow the coverage from different angles.
- arXiv · Computation and LanguageMARI:面向 LLM 推荐的蒸馏结构化决策记忆
论文提出 MARI(Memory-Augmented Recommendation with Interpretability),用离线 Post-Hoc Decision Distillation 把异构行为与用户内容蒸馏成结构化决策记忆(SDM),存入 Decision Memory Bank(DMB),在线检索相关 SDM 增强 LLM 推理。
Heat trend
Current heat 9·Comparable peak 10(Oct 9)·Comparable change over 24 hours –
The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.