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, and achieves over 10x higher online inference throughput.
Abstract
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.
Large Language Models (LLMs) have shown great promise in recommendation due to superior reasoning abilities. However, existing methods mainly rely on explicit Chain-of-Thought (CoT), resulting in verbose reasoning texts and inefficient response times. latent reasoning aims to balance efficiency by thinking within a continuous latent space, yet it faces two major challenges: (1) Lack of Fine-grained Supervision: Latent reasoning relies solely on feedback from the final labels, providing sparse supervisory signals that struggle to effectively guide the optimization of multiple hidden reasoning steps. (2) Single Reasoning Path: The deterministic nature of latent reasoning impedes the exploration of users'diverse interests and preferences, thereby limiting the recommendation capabilities of LLMs. To address these issues, we propose \textbf{$LaRec$}, an efficient generative recommendation framework designed to unleash the potential of latent reasoning in LLMs. $LaRec$ consists of two core stages: First, we design Latent Pre-training that empowers LLMs with latent reasoning capabilities by providing rich supervisory signals to the latent space reasoning via step-level alignment and process direction alignment. Second, we introduce Personalized RL-tuning. Specifically, we construct a personalized Gaussian Mixture Distribution for each user based on their historical interests. By randomly sampling distinct reasoning starting points from this distribution during training, we guide the LLMs to traverse diverse reasoning paths within the latent space, enabling efficient exploration of user's multi-faceted interests. Experiments on multiple datasets show that $LaRec$ significantly outperforms existing baselines with comparable efficiency.
Large Language Models (LLMs) have shown strong potential for recommendation by leveraging their semantic understanding and contextual modeling capabilities. Recent studies further introduce reasoning mechanisms to improve user preference modeling. However, explicit natural-language reasoning incurs substantial inference overhead, whereas existing latent reasoning methods mainly focus on generating or verifying intermediate states, leaving their layer-wise preference roles and contributions insufficiently characterized. We propose HiLaR, a Hierarchical Latent Reasoning framework with layer-aware reinforcement optimization for LLM-based recommendation. HiLaR constructs temporal-guided hierarchical user preference representations, aligns them with multiple LLM latent reasoning states, and organizes the reasoning process from broad preferences to fine-grained current intents. To further optimize the reasoning trajectory, HiLaR combines final recommendation feedback with layer-aware process rewards derived from the marginal target-likelihood gain of each state. Experiments on four Amazon benchmark datasets show that HiLaR generally outperforms strong sequential, generative, and LLM-based recommendation baselines. Ablation and sensitivity analyses further verify the contribution of hierarchical representation learning, latent alignment, and process-level optimization. Our code is available in https://github.com/hupeiyu21/HiLaR.
Large Language Models (LLMs) demonstrate significant potential in sequential recommendation, and leveraging their Chain-of-Thought (CoT) reasoning capabilities can further unlock profound user preference understanding. However, deploying explicit CoT reasoning in real-world systems faces prohibitive challenges: (i) the conflict between the large model scale required for high-fidelity reasoning and the resource constraints of online services, and (ii) the excessive latency introduced by auto-regressive rationale generation. To address these issues, we propose I Reasoning via Multi-Teacher Distillation (IRMD), a novel framework that 'compiles' the reasoning abilities of large teacher LLMs into a lightweight student Small Language Model (SLM). IRMD first employs a Multi-Teacher CoT Synthesis with Dual-Constraint Rejection Sampling module to generate a high-quality, diverse set of reasoning paths. Subsequently, our Annealing-Scheduled Reasoning Distillation strategy progressively trains the student to internalize this logic, transitioning from mimicking explicit CoT to performing purely implicit reasoning. Extensive experiments on multiple benchmark datasets demonstrate that IRMD significantly outperforms state-of-the-art baselines in both recommendation accuracy and inference efficiency. Our code is accessible at https://github.com/Cxx-0/IRMD.
Large Language Models (LLMs) have shown remarkable potential in recommendation systems but suffer from prohibitive inference latency. Existing distillation approaches typically target Small Language Models (SLMs) or Conventional Recommendation Models (CRMs), yet face a critical trade-off between computational cost and semantic reasoning capacity. To bridge this accuracy-efficiency gap, we introduce Reasoning-to-Encoder Distilla-tion (R2END), a framework that establishes a text encoder as the optimal student architecture for scalable recommendation. Unlike meth-ods that mimic token generation, R2END compresses the teacher’s reasoning into a dense vector space via a semantic alignment objec-tive, effectively capturing user-item dynamics. Extensive experiments on four datasets demonstrate that R2END not only outperforms state-of-the-art baselines but also achieves drastically reduced latency, offering a sweet spot for recommendation.
D. Han, Daeyoung Roh, A. Y. Kim et al.· Annual Meeting of the Associ...· 0 citations
The emergence of Large Language Models (LLMs) has driven a paradigm shift in sequential recommendation from discriminative ranking to generative modeling. However, existing generative approaches predominantly rely on semantic IDs (SIDs)—discrete identifiers derived from hierarchical quantization that function as semantic black boxes. Solely designed for in-domain reconstruction, these opaque tokens preclude the mining of explicit user behavioral logic. In contrast, plain-text modeling unlocks the full inferential potential of LLMs, enabling superior generalization and precise, logic-driven interest discovery. To bridge this gap, we propose GUIDER (Generative User Interest Discovery & Explicit Reasoning), a framework that fundamentally decouples intent planning from item matching by reformulating sequential modeling within a rigorous closed-set semantic interest space. GUIDER first internalizes domain-specific semantics via continual pre-training and activates explicit reasoning through a hindsight-guided chain-of-thought strategy. Crucially, to ensure precise optimization of the generated interest set, we introduce Token-Adaptive Policy Optimization (TAPO). Addressing the credit assignment problem—where standard reinforcement learning indiscriminately assigns identical advantages to the entire response regardless of local correctness—TAPO employs a token-level advantage adaptation mechanism. This strategy explicitly differentiates valid interest circles from hallucinations by dynamically redistributing advantages based on fine-grained quality, while a dynamic thinking penalty suppresses redundant reasoning. Extensive offline evaluations and large-scale online A/B testing on the Kuaishou platform validate that GUIDER not only achieves superior accuracy but also yields consistent gains in average app duration and effectively facilitates out-of-cocoon interest exploration.
Jinke Wu, Yinghao Wu, Shuchang Liu et al.· Proceedings of the 32nd ACM...· 0 citations
Large language model-based recommender systems (LLM-RSs) have demonstrated remarkable capabilities, but are computationally unsustainable for many real-world applications. Compact LLMs offer a practical alternative, yet their reduced capacity often requires reasoning or knowledge distillation methods that increase latency or depend on larger models. Combined with autoregressive generation, these approaches face severe scalability bottlenecks. In contrast, discriminative LLM-RSs enable efficient full-corpus ranking through embedding similarity, but compact backbones remain limited in expressiveness and structural adaptivity. We propose the Fusion of Layer-wise Exits for Sequential Recommendation (FLEXRec), a discriminative framework that enhances compact LLMs while retaining scalable full-corpus ranking. FLEXRec inserts prediction heads (i.e., exits) at multiple transformer layers and adaptively fuses their score distributions. An adaptive continuous router (AC-Router) dynamically selects both the number and identity of exits for each user sequence, while a novel target-k hinge loss regulates routing sparsity. Experiments on three real-world datasets with Qwen 3 1.7B and Llama 3.2 3B show that FLEXRec achieves state-of-the-art accuracy among competing methods while remaining highly efficient. Code: https://github.com/xurong-liang/FLEXRec
Xurong Liang, Tong Chen, Q. Nguyen et al.· 0 citations