arXiv:2512. 11529v3 Announce Type: replace Abstract: Recommendation system delivers substantial economic benefits by providing personalized predictions.
arXiv:2608. 09685v1 Announce Type: new Abstract: Autoregressive semantic ID recommenders are constrained by expensive beam-search decoding, which limits the practical length of item identifiers.
arXiv:2602. 22647v2 Announce Type: replace-cross Abstract: Generative retrieval has emerged as a powerful paradigm for LLM-based recommendation.
arXiv:2607. 28659v1 Announce Type: new Abstract: Cross-domain sequential recommendation (CDSR) aims to model users' dynamic interest transitions and sequential patterns across multiple domains.
Orthrus is a serving system that performs heterogeneous batching of embedding and generative models within a single inference loop. It uses chunked embedding with incremental pooling and workload‑aware batch composition to unify conflicting computational patterns. Experiments on four A100 GPUs show that Orthrus improves throughput by 1.28×–4.52× and reduces p99 latency by up to 55.8% compared to baseline deployments.
arXiv:2604. 24806v2 Announce Type: replace-cross Abstract: Modern Deep Learning Recommendation Models (DLRMs) follow scaling laws with sequence length, driving the frontier toward ultra-long User Interaction History (UIH).
arXiv:2607. 24804v2 Announce Type: replace-cross Abstract: Recommendation systems have undergone significant transformations in the past years.
DeGRe is a dense‑supervised generative reranking framework designed to improve multi‑stage recommender systems by addressing label bias and credit assignment issues. It uses an offline Lookahead Evaluator with beam search to generate dense supervision signals, which are distilled into a lightweight Online Generator that can perform efficient greedy decoding at inference time. Experiments show that DeGRe outperforms baselines on public benchmarks and industrial datasets, and it has been successfully deployed on Taobao Flash Shopping to enhance online recommendations.
arXiv:2607. 24845v1 Announce Type: cross Abstract: Large language models (LLMs) have been applied to sequential recommendation by formulating it as a natural language task.
SequenceO1 is an end‑to‑end framework that enables ultra‑long (up to 100K interactions) sequence modeling for recommendation systems. It compresses raw user histories into a fixed‑size sketch using Sketch Attention and then models short‑term and long‑term interests with Target‑to‑History Cross Attention. The system incorporates low‑rank caching, batching, pipeline lift, and a FlashSA kernel to keep training and inference efficient, achieving consistent offline and online performance gains when deployed at full traffic on Douyin.
arXiv:2606. 11023v1 Announce Type: cross Abstract: Sequential recommendation aims to predict users' next interaction with items by analyzing their historical behavior.
arXiv:2607. 13095v1 Announce Type: cross Abstract: We present a full-pipeline inference optimization for the MiMo-V2.
arXiv:2606. 07546v1 Announce Type: cross Abstract: Capturing user interests across extensive watch histories is critical for short-form video recommendation, yet scaling sequence length is limited by two bottlenecks: the semantic sparsity of atomic Video IDs and the quadratic computational complexity of Transformers.