跳到正文
arXiv cs.AI· Haoyi Hu, Gaoyang Guo, Yunjia Xi, Wen Chen, Weiwen Liu, Jian Wu, Yuning Jiang, Weinan Zhang·· 4 小时前AI 评分27

RecMem:面向长序列生成式推荐的循环记忆压缩框架

RecMem: Recurrent Memory Compression for Long-Sequence Generative Recommendation

AI 导读

论文提出 RecMem,一种循环记忆压缩框架,将历史交互压缩为固定记忆槽位,使记忆状态与解码器上下文长度不受历史长度影响。

正文 · 原文

View PDF HTML (experimental)

Abstract:Generative recommendation benefits from long interaction histories, but the cost of processing them directly grows with history length. We propose RecMem, a recurrent memory compression framework guided by three desiderata: space invariance, semantic fidelity, and lifelong evolvability. RecMem compresses historical segments into a fixed set of memory slots using a bifurcated encoder-decoder with shared lower layers and slot-wise forget-gated updates. Recommendation supervision trains the memory to preserve predictive information, while stream-consistency training aligns predictions from sequential and packed processing. By construction, the memory state and decoder context are independent of history length; under contraction and approximate semigroup consistency, we further derive a horizon-independent bound on the state discrepancy between these two processing orders. On MerRec, RecMem uses 16 memory slots and a 216-token decoder context, compared with 1,200 tokens for HSTU-1200, while retaining 87.9% of HSTU-1200's Recall@10 and 97.0% of its Recall@200. It outperforms PHSTU and the evaluated recent-only and retrieval baselines, and analyses of distant history show that archived interactions contribute predictive signal.
Comments: Accepted at NeurIPS 2026
Subjects: Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
Cite as: arXiv:2611.17921 [cs.AI]
  (or arXiv:2611.17921v1 [cs.AI] for this version)
  https://doi.org/10.48550/arXiv.2611.17921

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Gaoyang Guo [view email]
[v1] Fri, 13 Nov 2026 08:59:44 UTC (267 KB)

来源:arXiv cs.AI · arxiv.org