cs.AINov 13, 2026

RecMem: Recurrent Memory Compression for Long-Sequence Generative Recommendation

Authors: Haoyi Hu, Gaoyang Guo, Yunjia Xi, Wen Chen, Weiwen Liu, Jian Wu, Yuning Jiang, Weinan Zhang

Organizations: Shanghai Jiao Tong University · Taobao & Tmall Group of Alibaba

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.

Figures & tables

Appendix figures & tables14 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Is Sliding Window All You Need? An Open Framework for Long-Sequence Recommendation

    Apr 14, 2026Sayak Chakrabarty, Souradip PalSequential RecommendationSequence Modeling

  2. MARS: Multi-resolution Adaptive Routing for Sequential Recommendation

    Oct 5, 2026Ming Yin, Sixun Dong, Yudong Liu +3Sequential RecommendationUser Modeling