Paper

SinkRec: Mitigating Semantic State Sink in Long Sequence Recommendation with Memory-Conditioned Gated Delta Networks

Unreadunread

AI summary

arXiv:2606. 09888v1 Announce Type: new Abstract: Linear attention provides an efficient backbone for long-sequence recommendation by avoiding the quadratic cost of standard Transformers, but its compressed recurrent state can be dominated by repetitive behavior patterns.

Reading status

Priority

About this paper

Authors: Zhuang Zhuang, Zhipeng Wei, Ji Dai, Jie Chen, Fei Pan, Peng Jiang, Kun Gai
Released: 2026-06-10 (2mo ago)
Topic: Related to current project
Relation to my work: Read next because SinkRec: Mitigating Semantic State Sink in Long Sequence Recommendation with Memory-Conditioned Gated Delta Networks overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, line, rate, model. Source: arxiv cs.LG (Machine Learning).
arXiv: 2606.09888

Ask Claude about this paper

Posts a comment with @claude so the runner spins up a Claude Code session scoped to SinkRec: Mitigating Semantic State Sink in Long Sequence Recommendation with Memory-Conditioned Gated Delta Networks. Replies appear in the conversation below.

Comments

No comments yet.

Related records

No links yet.