Do Recommenders Need to Think in Words? Latent Reasoning for Efficient LLM-Based Recommendation
cs.CL
Submitted: 2026-11-05
Updated: 2026-11-05
Comments: Accepted at CIKM 2026 (Short Paper). 5 pages, 2 figures, 3 tables
Journal ref: Proceedings of the 35th ACM International Conference on Information and Knowledge Management (CIKM '26), November 7-11, 2026, Rome, Italy
License: http://creativecommons.org/licenses/by/4.0/
The gist: LLM-based recommender systems increasingly generate chain-of-thought (CoT) text before each item prediction to activate the LLM's reasoning capacity…
Terminology
Abstract
LLM-based recommender systems increasingly generate chain-of-thought (CoT) text before each item prediction to activate the LLM's reasoning capacity, costing tens to hundreds of extra tokens per recommendation at inference. We ask whether inference-time reasoning artifacts (textual or latent) can be replaced by training-time supervision alone, without loss of accuracy. We propose LatentRec, a framework for LLM-based generative recommendation that replaces inference-time reasoning artifacts with training-time per-step head supervision. LatentRec has three components. A 2.5M-parameter Continuous Thought Module iterates K latent reasoning steps over the LLM's hidden state during training. Soft Token Selection bridges the 14x norm gap between hidden states and input embeddings. A Per-step Reasoning Loss (PRL) supervises every intermediate state through the LLM's prediction head against the first token of the target Semantic ID. The trained LLM retains this accuracy without the module at inference, so the module can be discarded at deployment. On three Amazon benchmarks, LatentRec improves HR@10 by 4.0-5.0% (relative) over a CoT-distilled baseline at 18.3x lower latency (batch=1). Per-step training supervision consistently improves accuracy across three datasets, paid once at training rather than at every recommendation.
Sources
- Training Large Language Models to Reason in a Continuous Latent Space
- Session-based Recommendations with Recurrent Neural Networks
- MiniOneRec: An Open-Source Framework for Scaling Generative Recommendation
- LARES: Latent Reasoning for Sequential Recommendation
- OneRec-Think: In-Text Reasoning for Generative Recommendation
- Parallel Latent Reasoning for Sequential Recommendation
- Qwen3 Technical Report
- Reinforced Latent Reasoning for LLM-based Recommendation
Related papers
- ALARM: Audio-Language Alignment for Reasoning Models
- ThinkSum: Probabilistic reasoning over sets using large language models
- Coherence boosting: When your pretrained language model is not paying enough attention
- From Facts to Insights: A Persona-Driven Dual Memory Framework and Dataset for Role-Playing Agents
- Improving General Role-Playing Agents via Psychology-Grounded Reasoning and Role-Aware Policy Optimization
- The Art of Asking: Multilingual Prompt Optimization for Synthetic Data