2 papers
cs.IR2026
GateDiffInt: Gate-Mediated Controllable Diffusion and Multi-Intent LLM Distillation for User Behavior Modeling
Jialong Duan, Zichen Zhang, Zirui Tu +7
Existing ranking models encode intent only implicitly, making it hard to disentangle structured intents of varying strength and temporal scale. Noise and intent in behavior sequenc…
cs.AI2026
Mamba with Hierarchical Memory: Solving Representation Bottleneck in Long Sequence Modeling
Qinwen Wang, Jieping Luo, Aoxiang Qin +5
Recurrent linear attention models (RLAs) such as Mamba offer efficient linear-time sequence modeling as an alternative to Transformers, yet their fixed-capacity recurrent states li…