Implicit Reasoning for Large Language Model-based Generative Recommendation

2026-06-12

AI Summary

This work addresses the limitations of large language models (LLMs) in generative recommendation, where incompatibility between semantic IDs and natural language reasoning interfaces hinders performance, and existing explicit reasoning approaches suffer from high training costs and marginal gains. To overcome these challenges, the authors propose PauseRec—a lightweight implicit reasoning paradigm that bypasses explicit chain-of-thought training and instead directly elicits the LLM’s intrinsic reasoning capabilities. By aligning semantic ID representations with natural language embedding spaces and employing end-to-end lightweight training, PauseRec achieves substantial improvements in both efficiency and effectiveness: it boosts recommendation performance by up to 6.22% over explicit chain-of-thought methods, reduces training cost by 65 GPU hours, and accelerates inference speed by 71.3%.

Abstract

Large Language Models (LLMs) are increasingly adopted as backbones for Generative Recommendation (GR), promising access to pretrained world knowledge. Yet reliably invoking this knowledge for GR remains poorly understood. A key obstacle is that LLM-based GR typically represents items with Semantic IDs (SIDs), disrupting LLMs' natural-language reasoning interface because these tokens are unseen by the LLM during pretraining. Existing approaches address this with expensive multi-stage pipelines that ground SIDs and elicit explicit rationales, but offer limited insight into when and why each stage is necessary. In this work, we systematically decompose explicit reasoning training pipelines for LLM-based GR, revealing three key limitations: weakened world-knowledge verbalization, misalignment between SID and natural-language token embedding spaces, and sensitivity to rationale quality, all of which hurt explicit reasoning performance. To circumvent these issues, we propose PauseRec, a lightweight implicit reasoning paradigm tailored for GR. PauseRec is exceptionally practical, avoiding costly reasoning trace acquisition and reasoning alignment training, leading to a multitude of benefits: (1) it outperforms standard explicit CoT methods by up to 6.22%, (2) it reduces training cost by up to 65% GPU hours, and (3) it speeds up inference by up to 71.3%. These results position PauseRec as a lightweight alternative to explicit rationale generation, enabling more effective and efficient LLM-based GR.

Problem

• Large Language Models
• Generative Recommendation
• Semantic IDs
• Implicit Reasoning
• World Knowledge

Innovation

• Implicit Reasoning
• Generative Recommendation
• Large Language Models
• Semantic IDs
• PauseRec