Searcharxiv⌕ Search

arXiv subjects

Kuo Cai

Publications and source records attributed to Kuo Cai.

At least 19 recordsLinked to original sources
arXiv2026

MuonRec: Shifting the Optimizer Paradigm Beyond Adam in Scalable Generative Recommendation

Recommender systems (RecSys) are increasingly emphasizing scaling, leveraging larger architectures and more interaction data to improve personalization. Yet, despite the optimizer's pivotal role in training, modern RecSys pipelines almost universally default to Adam/AdamW, with limited scrutiny of whether these choices are truly optimal for recommendation. In this work, we revisit optimizer design for scalable recommendation and introduce MuonRec, the first framework that brings the recently proposed Muon optimizer to RecSys training. Muon performs orthogonalized momentum updates for 2D weight matrices via Newton-Schulz iteration, promoting diverse update directions and improving optimization efficiency. We develop an open-source training recipe for recommendation models and evaluate it across both traditional sequential recommenders and modern generative recommenders. Extensive experiments demonstrate that MuonRec reduces converged training steps by an average of 32.4\% while simultaneously improving final ranking quality. Specifically, MuonRec yields consistent relative gains in NDCG@10, averaging 12.6\% across all settings, with particularly pronounced improvements in generative recommendation models. These results consistently outperform strong Adam/AdamW baselines, positioning Muon as a promising new optimizer standard for RecSys training. Our code is available.

cs.IR↗
arXiv2026

OneLatent: Latent Reasoning for Efficient Foundation Recommendation Models

Large language models (LLMs) have demonstrated strong reasoning capabilities, motivating their use as the backbone of foundation recommendation models (FRMs). Existing methods enhance recommendations through explicit Chain-of-Thought (CoT) reasoning under a Think-then-Answer paradigm. However, explicit CoT incurs substantial inference overhead by generating lengthy reasoning traces and relies on manually designed templates that struggle to capture diverse, dynamic user interests. We propose OneLatent, an efficient latent reasoning framework that compresses explicit reasoning traces into several learnable latent tokens, enabling Latent-Reason-then-Answer inference without generating verbose traces. OneLatent first introduces Multi-View Adaptive CoT (MV-ACoT), which creates diverse, high-quality teacher-generated supervision by exploring user interests from multiple perspectives and automatically adapting reasoning complexity to each instance. Building on pretrained FRMs, it then uses a three-stage latent-token alignment paradigm to progressively internalize CoT traces into learnable latent tokens. Finally, a multistage curriculum-based post-training strategy activates latent-token reasoning for downstream recommendation tasks. Experiments on an industrial-scale Kuaishou dataset and the public Kuaishou LLM-Rec benchmark show that OneLatent consistently outperforms explicit CoT-based methods and traditional baselines. Compared with the Think and No-Think variants of FRMs, OneLatent improves SID@64 by 17.44% and 9.33%, respectively, while achieving over 17x higher online inference throughput. We further develop a production serving system for scalable, real-time FRM inference. An online A/B test in Kuaishou's local-services advertising scenario shows that deploying OneLatent with this system yields an estimated 9.6% revenue lift over strong online baselines, including OneRec and OneReason.

cs.IR↗
arXiv2026

Utility-Aware Generative Recommendation: Utility-Enriched Semantic Codes and Utility-Guided Decoding

Generative recommendation has been widely adopted in large-scale industrial recommendation systems, where multimodal item embeddings are quantized into offline semantic identifiers (SIDs) and items are retrieved by autoregressively generating their SIDs. However, conventional SID construction often underrepresents item utility reflected in post-exposure behaviors such as shares, long views, and video completion, making semantically similar items with different utility insufficiently distinguishable while struggling to accommodate rapidly changing utility signals. Moreover, generation likelihood may be misaligned with downstream utility, causing high-utility candidates to be prematurely pruned during beam search before post-generation reranking or training-time reward alignment can effectively compensate. To address these limitations, we propose UAGR, which aligns both SID construction and decoding with downstream multi-objective utility. Our Utility-Enriched Semantic Codes preserve the original semantic information while incorporating item-level utility without increasing SID length or vocabulary size, supporting incremental utility updates with no additional tokenization or sequence-decoding cost. Meanwhile, our Utility-Guided Decoding estimates candidate utility by combining predictions of multiple post-exposure outcomes and incorporates these estimates into beam scoring, reducing the risk of prematurely pruning high-utility paths while enabling flexible utility control without interfering with generative backbone training. Extensive offline experiments and online A/B tests under real-world traffic demonstrate the effectiveness and deployability of UAGR, which has been fully deployed on Kuaishou and increases video watch time by 0.206\% online.

cs.LG↗
arXiv2026

WhisperRec: Latent Reasoning for Efficient Foundation Recommendation Models

Large language models (LLMs) have demonstrated strong reasoning capabilities, motivating their adoption as backbones for foundation recommendation models (FRMs). Existing approaches typically enhance recommendation with explicit Chain-of-Thought (CoT) under the Think-then-Answer paradigm. However, generating lengthy rationales introduces substantial inference overhead, while fixed CoT templates struggle to model diverse, dynamic, and context-dependent user interests. We propose WhisperRec, an efficient latent reasoning framework for FRMs. WhisperRec compresses teacher-generated CoT into learnable latent reasoning tokens, enabling a Latent-Reason-then-Answer paradigm that performs reasoning in latent space without producing verbose rationales. This design retains decision-relevant reasoning information while avoiding the latency bottleneck of autoregressive rationale generation. Specifically, it first introduces Multi-View Adaptive CoT (MV-ACoT) to construct diverse, high-quality supervision from complementary perspectives on user interests. MV-ACoT also adapts reasoning complexity to each instance, applying lightweight analysis to clear cases and targeted multi-factor reasoning to challenging ones. Building on a pre-trained FRM, WhisperRec then employs a three-stage Latent Reasoning Alignment procedure to progressively internalize teacher CoT into latent representations. Finally, curriculum-based post-training activates latent-token reasoning for downstream recommendation while preserving standard recommendation capability. Experiments on an industrial-scale Kuaishou dataset and the public Kuaishou LLM-Rec benchmark show that WhisperRec consistently outperforms explicit-CoT methods and conventional baselines. Compared with explicit CoT Think and No-Think variants, WhisperRec improves SID@64 by 17.44% and 9.33%, respectively, and achieves over 10x higher online inference throughput.

cs.IR↗
arXiv2026

Kwai Summary Attention V2 Technical Report

Summary-token architectures such as Kwai Summary Attention (KSA) compress distant context into learnable Summary token per chunk, and hybrid variants interleave these layers with full attention to retain exact access to the original text. In such hybrid KSA-Full architecture, the interleaving inserted summary tokens are served as a streaming flow role across different layers. Although effective, however, it also introduce some bottlenecks: (1) the different layers summary compression chunk must be same, therefore hard to reach a flexible KV Cache compression, e.g., different layer set different summary compression chunk window. (2) in the full-attention layers, the summary token introduced additional computation cost, but such layer ignore them and still pay quadratic prefilling, therefore unfriendly to the computation side. Motivated by them, we developed the KSA V2, which not only achieves a more flexible KV cache memory compression but also reduces the computation cost in Full attention layer. Specifically, we found that the summary tokens can serve two roles: the KV memory and sparse indexer.

cs.LG↗
arXiv2026

CRM: Retrieval Model with Controllable Condition

Retrieval models for recommendation need to account for outcomes such as watch time while searching large item corpora efficiently. We present the Controllable Retrieval Model (CRM), an outcome-conditioned retrieval framework that adds watch time as an input to the user tower. During training, CRM learns to predict the next item given its observed watch time; at inference, we supply a desired watch time to steer retrieval without retraining. The item tower is independent of the condition, so item embeddings can still be precomputed and indexed for approximate nearest neighbor (ANN) search. We implement CRM as both a DNN-based two-tower model and a Decision Transformer. The Transformer interleaves item tokens with watch-time-to-go condition tokens, using the target watch time as a numerical prompt for next-item prediction. For each online request, we choose between the average and maximum watch times in the user's recent history as the condition. CRM is deployed in Kuaishou's short-video recommendation system, which serves over 400 million users. Online A/B tests show video watch time improvements of 0.372% in Kuaishou and 0.457% in Kuaishou Lite.

cs.IR↗
arXiv2026

EvoSkillRec: Skill-Genome Evolution for Recommender Architecture Discovery

Modern recommender systems advance not only by scaling data and parameters, but also by encoding task-specific inductive biases through architecture, including sparse feature interactions for click-through rate (CTR) prediction, temporal attention for sequential recommendation, and expert routing for multi-task learning. However, these biases are typically human expert designed or searched within predefined operator spaces. Although Recent LLM-driven code evolution expands this space, unconstrained edits often produce invalid or ineffective architectures, underuse established architecture design knowledge, and fail to preserve successful innovations for reuse. We introduce EvoSkillRec, a promotion-and-reuse framework for cumulative recommender architecture evolution. It first decomposes recommenders into atomic executable skills and represents architectures as typed skill genomes, with each skill equipped with input--output types, semantic annotations, and implementation code. We then evolve models with different tasks through two coupled spaces: a constrained skill--space that mutates, recombines, specializes, and reuses validated skills, and an open-ended code--space in which LLM planners and synthesizers invent new skill modules using prior evolution traces and accumulated experience. An autoresearch controller evaluates candidates, diagnoses failures, retrieves relevant skills, promotes validated innovations into the skill library, and adaptively allocates the proposal budget between the two spaces. Extensive experiments on CTR prediction, multi-task learning, and multi-domain learning, including resource-constrained co-optimization of predictive quality and model FLOPs utilization in generative ranking models, consistently demonstrate the effectiveness of our proposed EvoSkillRec.

cs.IR↗
arXiv2026

Hierarchical Quantization with Domain-Adaptive Sparse Routing for Generative Cross-Domain Recommendation

Generative Recommendation (GenRec) represents a promising paradigm that achieves remarkable empirical success by encoding items as compact Semantic IDs (SIDs) and modeling user behavior via next-token prediction across diverse recommendation scenarios. Extending this paradigm to cross-domain recommendation is challenging because a unified model must accommodate heterogeneous item semantics and behavioral patterns across domains. Existing methods commonly rely on globally shared representations or lightweight domain adaptation, which may provide insufficient capacity for modeling heterogeneous patterns at different semantic granularities. To address these challenges, we propose HD-Rec, a unified generative framework for cross-domain recommendation. HD-Rec employs a hierarchical domain-aware quantizer that constructs semantic identifiers using globally shared coarse-level codebooks and adaptively routed fine-level codebooks. It further introduces a domain-adaptive sparse mixture-of-experts module that combines a continuously activated shared expert with a dynamically selected specialized expert. To improve the coherence of multi-token item representations, we develop a cross-granularity routing consistency objective that regularizes token-level routing decisions toward their item-level consensus. Experiments on three public cross-domain recommendation benchmarks show that HD-Rec consistently improves over competitive sequential, generative, and cross-domain recommendation baselines.

cs.IR↗
arXiv2026

Coarse-to-Fine Long-term Interest Modeling for Generative Recommendation

Leveraging long-term user behavioral patterns is a key trajectory for enhancing the accuracy of modern recommender systems. While generative recommender systems have emerged as a transformative paradigm, they face hurdles in effectively modeling extensive historical sequences. To address this challenge, we propose GLASS, a novel framework that integrates long-term user interests into the generative process via SID-Tier and Semantic Search. We first introduce SID-Tier, a module that maps long-term interactions into a unified interest vector to enhance the prediction of the initial SID token. Unlike traditional retrieval models that struggle with massive item spaces, SID-Tier leverages the compact nature of the semantic codebook to incorporate cross features between the user's long-term history and candidate semantic codes. Furthermore, we present semantic hard search, which utilizes generated coarse-grained semantic ID as dynamic keys to extract relevant historical behaviors, which are then fused via an adaptive gated fusion module to recalibrate the trajectory of subsequent fine-grained tokens. To address the inherent data sparsity in semantic hard search, we propose two strategies: semantic neighbor augmentation and codebook resizing. Extensive experiments on two large-scale real-world datasets, TAOBAO-MM and KuaiRec, demonstrate that GLASS outperforms state-of-the-art baselines, achieving significant gains in recommendation quality. Our codes are made publicly available to facilitate further research in generative recommendation.

cs.IR↗
arXiv2026

OneReason Technical Report

Generative recommendation models in the OneRec family have been widely deployed in many real-world services, such as short-video, live-streaming, advertising, and e-commerce. However, these generative models can only benefit from the scaling advantage, while their reasoning ability is hard to activate, since we cannot construct meaningful Chain-of-Thought (CoT) sequences consisting of itemic tokens only. Inspired by the success of the reasoning-style ``think before answer'' paradigm in the LLM field, we conduct preliminary studies (i.e., OneRec-Think, OpenOneRec) to explore reasoning capability in generative recommendation. Nevertheless, we notice an unexpected phenomenon: the thinking mode does not show advantages over the non-thinking mode. Drawing insights from recent findings on CoT robustness in multi-modal language models, we argue that effective reasoning in recommendation rests on two factors: perception, the ability to ground itemic tokens in their underlying language semantics, and cognition, the ability to reorganize a user's behavior sequence into coherent latent interest points. We therefore propose OneReason, which includes: (1) strong itemic token perception in pre-training, (2) a three-level cognition-enhanced CoT format for recommendation tasks in SFT, and (3) a specialize-then-unify training recipe in RL to enhance the thinking ability.

cs.IR↗
arXiv2026

GEMs: Breaking the Long-Sequence Barrier in Generative Recommendation with a Multi-Stream Decoder

While generative recommendations (GR) possess strong sequential reasoning capabilities, they face significant challenges when processing extremely long user behavior sequences: the high computational cost forces practical sequence lengths to be limited, preventing models from capturing users' lifelong interests; meanwhile, the inherent "recency bias" of attention mechanisms further weakens learning from long-term history. To overcome this bottleneck, we propose GEMs (Generative rEcommendation with a Multi-stream decoder), a novel and unified framework designed to break the long-sequence barrier by capturing users' lifelong interaction sequences through a multi-stream perspective. Specifically, GEMs partitions user behaviors into three temporal streams$\unicode{x2014}$Recent, Mid-term, and Lifecycle$\unicode{x2014}$and employs tailored inference schemes for each: a one-stage real-time extractor for immediate dynamics, a lightweight indexer for cross attention to balance accuracy and cost for mid-term sequences, and a two-stage offline-online compression module for lifelong modeling. These streams are integrated via a parameter-free fusion strategy to enable holistic interest representation. Extensive experiments on large-scale industrial datasets demonstrate that GEMs significantly outperforms state-of-the-art methods in recommendation accuracy. Notably, GEMs is the first lifelong GR framework successfully deployed in a high-concurrency industrial environment, achieving superior inference efficiency while processing user sequences of over 100,000 interactions.

cs.IR↗
arXiv2026

OneLive: Dynamically Unified Generative Framework for Live-Streaming Recommendation

Live-streaming recommender system serves as critical infrastructure that bridges the patterns of real-time interactions between users and authors. Similar to traditional industrial recommender systems, live-streaming recommendation also relies on cascade architectures to support large-scale concurrency. Recent advances in generative recommendation unify the multi-stage recommendation process with Transformer-based architectures, offering improved scalability and higher computational efficiency. However, the inherent complexity of live-streaming prevents the direct transfer of these methods to live-streaming scenario, where continuously evolving content, limited lifecycles, strict real-time constraints, and heterogeneous multi-objectives introduce unique challenges that invalidate static tokenization and conventional model framework. To address these issues, we propose OneLive, a dynamically unified generative recommendation framework tailored for live-streaming scenario. OneLive integrates four key components: (i) A Dynamic Tokenizer that continuously encodes evolving real-time live content fused with behavior signal through residual quantization; (ii) A Time-Aware Gated Attention mechanism that explicitly models temporal dynamics for timely decision making; (iii) An efficient decoder-only generative architecture enhanced with Sequential MTP and QK Norm for stable training and accelerated inference; (iv) A Unified Multi-Objective Alignment Framework reinforces policy optimization for personalized preferences.

cs.IR↗
arXiv2026

OpenOneRec Technical Report

While the OneRec series has successfully unified the fragmented recommendation pipeline into an end-to-end generative framework, a significant gap remains between recommendation systems and general intelligence. Constrained by isolated data, they operate as domain specialists-proficient in pattern matching but lacking world knowledge, reasoning capabilities, and instruction following. This limitation is further compounded by the lack of a holistic benchmark to evaluate such integrated capabilities. To address this, our contributions are: 1) RecIF Bench & Open Data: We propose RecIF-Bench, a holistic benchmark covering 8 diverse tasks that thoroughly evaluate capabilities from fundamental prediction to complex reasoning. Concurrently, we release a massive training dataset comprising 96 million interactions from 160,000 users to facilitate reproducible research. 2) Framework & Scaling: To ensure full reproducibility, we open-source our comprehensive training pipeline, encompassing data processing, co-pretraining, and post-training. Leveraging this framework, we demonstrate that recommendation capabilities can scale predictably while mitigating catastrophic forgetting of general knowledge. 3) OneRec-Foundation: We release OneRec Foundation (1.7B and 8B), a family of models establishing new state-of-the-art (SOTA) results across all tasks in RecIF-Bench. Furthermore, when transferred to the Amazon benchmark, our models surpass the strongest baselines with an average 26.8% improvement in Recall@10 across 10 diverse datasets (Figure 1). This work marks a step towards building truly intelligent recommender systems. Nonetheless, realizing this vision presents significant technical and theoretical challenges, highlighting the need for broader research engagement in this promising direction.

cs.IR↗
arXiv2026

ALPBench: A Benchmark for Attribution-level Long-term Personal Behavior Understanding

Recent advances in large language models have highlighted their potential for personalized recommendation, where accurately capturing user preferences remains a key challenge. Leveraging their strong reasoning and generalization capabilities, LLMs offer new opportunities for modeling long-term user behavior. To systematically evaluate this, we introduce ALPBench, a Benchmark for Attribution-level Long-term Personal Behavior Understanding. Unlike item-focused benchmarks, ALPBench predicts user-interested attribute combinations, enabling ground-truth evaluation even for newly introduced items. It models preferences from long-term historical behaviors rather than users' explicitly expressed requests, better reflecting enduring interests. User histories are represented as natural language sequences, allowing interpretable, reasoning-based personalization. ALPBench enables fine-grained evaluation of personalization by focusing on the prediction of attribute combinations task that remains highly challenging for current LLMs due to the need to capture complex interactions among multiple attributes and reason over long-term user behavior sequences.

cs.IR↗
arXiv2026

OneMall: One Architecture, More Scenarios -- End-to-End Generative Recommender Family at Kuaishou E-Commerce

In the wave of generative recommendation, we present OneMall, an end-to-end generative recommendation framework tailored for e-commerce services at Kuaishou. Our OneMall systematically unifies the e-commerce's multiple item distribution scenarios, such as Product-card, short-video and live-streaming. Specifically, it comprises three key components, aligning the entire model training pipeline to the LLM's pre-training/post-training: (1) E-commerce Semantic Tokenizer: we provide a tokenizer solution that captures both real-world semantics and business-specific item relations across different scenarios; (2) Transformer-based Architecture: we largely utilize Transformer as our model backbone, e.g., employing Query-Former for long sequence compression, Cross-Attention for multi-behavior sequence fusion, and Sparse MoE for scalable auto-regressive generation; (3) Reinforcement Learning Pipeline: we further connect retrieval and ranking models via RL, enabling the ranking model to serve as a reward signal for end-to-end policy retrieval model optimization. Extensive experiments demonstrate that OneMall achieves consistent improvements across all e-commerce scenarios: +13.01\% GMV in product-card, +15.32\% Orders in Short-Video, and +2.78\% Orders in Live-Streaming. OneMall has been deployed, serving over 400 million daily active users at Kuaishou.

cs.IR↗
arXiv2026

Unleashing the Native Recommendation Potential: LLM-Based Generative Recommendation via Structured Term Identifiers

Leveraging the vast open-world knowledge and understanding capabilities of Large Language Models (LLMs) to develop general-purpose, semantically-aware recommender systems has emerged as a pivotal research direction in generative recommendation. However, existing methods face bottlenecks in constructing item identifiers. Text-based methods introduce LLMs' vast output space, leading to hallucination, while methods based on Semantic IDs (SIDs) encounter a semantic gap between SIDs and LLMs' native vocabulary, requiring costly vocabulary expansion and alignment training. To address this, this paper introduces Term IDs (TIDs), defined as a set of semantically rich and standardized textual keywords, to serve as robust item identifiers. We propose GRLM, a novel framework centered on TIDs, employs Context-aware Term Generation to convert item's metadata into standardized TIDs and utilizes Integrative Instruction Fine-tuning to collaboratively optimize term internalization and sequential recommendation. Additionally, Elastic Identifier Grounding is designed for robust item mapping. Extensive experiments on real-world datasets demonstrate that GRLM significantly outperforms baselines across multiple scenarios, pointing a promising direction for generalizable and high-performance generative recommendation systems.

cs.IR↗
arXiv2026

PROMISE: Process Reward Models Unlock Test-Time Scaling Laws in Generative Recommendations

Generative Recommendation has emerged as a promising paradigm, reformulating recommendation as a sequence-to-sequence generation task over hierarchical Semantic IDs. However, existing methods suffer from a critical issue we term Semantic Drift, where errors in early, high-level tokens irreversibly divert the generation trajectory into irrelevant semantic subspaces. Inspired by Process Reward Models (PRMs) that enhance reasoning in Large Language Models, we propose Promise, a novel framework that integrates dense, step-by-step verification into generative models. Promise features a lightweight PRM to assess the quality of intermediate inference steps, coupled with a PRM-guided Beam Search strategy that leverages dense feedback to dynamically prune erroneous branches. Crucially, our approach unlocks Test-Time Scaling Laws for recommender systems: by increasing inference compute, smaller models can match or surpass larger models. Extensive offline experiments and online A/B tests on a large-scale platform demonstrate that Promise effectively mitigates Semantic Drift, significantly improving recommendation accuracy while enabling efficient deployment.

cs.IR↗
arXiv2025

OneRec-Think: In-Text Reasoning for Generative Recommendation

The powerful generative capacity of Large Language Models (LLMs) has instigated a paradigm shift in recommendation. However, existing generative models (e.g., OneRec) operate as implicit predictors, critically lacking the capacity for explicit and controllable reasoning-a key advantage of LLMs. To bridge this gap, we propose OneRec-Think, a unified framework that seamlessly integrates dialogue, reasoning, and personalized recommendation. OneRec-Think incorporates: (1) Itemic Alignment: cross-modal Item-Textual Alignment for semantic grounding; (2) Reasoning Activation: Reasoning Scaffolding to activate LLM reasoning within the recommendation context; and (3) Reasoning Enhancement, where we design a recommendation-specific reward function that accounts for the multi-validity nature of user preferences. Experiments across public benchmarks show state-of-the-art performance. Moreover, our proposed "Think-Ahead" architecture enables effective industrial deployment on Kuaishou, achieving a 0.159\% gain in APP Stay Time and validating the practical efficacy of the model's explicit reasoning capability.

cs.IR↗