Verifiable Reasoning for LLM-based Generative Recommendation

2026-03-08

AI Summary

This work addresses a critical limitation in existing large language model–based generative recommendation approaches, which typically follow a “reason-then-recommend” paradigm lacking intermediate validation and thus prone to reasoning degradation—such as homogenization or error accumulation—that undermines recommendation quality. To mitigate this, we propose a novel “reason–verify–recommend” paradigm that integrates verification directly into the reasoning process, enabling reliable feedback to guide more accurate user preference modeling. We introduce two core design principles—reliability and multidimensionality—and develop a hybrid verifier architecture optimized through proxy prediction objectives, resulting in VRec, a verifiable reasoning–based recommendation system. Extensive experiments on four real-world datasets demonstrate that VRec significantly improves both recommendation performance and scalability while maintaining high inference efficiency.

Technology Category

• Knowledge Representation and Reasoning: Reasoning with Beliefs
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Reasoning under Uncertainty: Other Foundations of Reasoning under Uncertainty

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Reasoning in Large Language Models (LLMs) has recently shown strong potential in enhancing generative recommendation through deep understanding of complex user preference. Existing approaches follow a {reason-then-recommend} paradigm, where LLMs perform step-by-step reasoning before item generation. However, this paradigm inevitably suffers from reasoning degradation (i.e., homogeneous or error-accumulated reasoning) due to the lack of intermediate verification, thus undermining the recommendation. To bridge this gap, we propose a novel \textbf{\textit{reason-verify-recommend}} paradigm, which interleaves reasoning with verification to provide reliable feedback, guiding the reasoning process toward more faithful user preference understanding. To enable effective verification, we establish two key principles for verifier design: 1) reliability ensures accurate evaluation of reasoning correctness and informative guidance generation; and 2) multi-dimensionality emphasizes comprehensive verification across multi-dimensional user preferences. Accordingly, we propose an effective implementation called VRec. It employs a mixture of verifiers to ensure multi-dimensionality, while leveraging a proxy prediction objective to pursue reliability. Experiments on four real-world datasets demonstrate that VRec substantially enhances recommendation effectiveness and scalability without compromising efficiency. The codes can be found at https://github.com/Linxyhaha/Verifiable-Rec.

Problem

• reasoning degradation
• generative recommendation
• Large Language Models
• intermediate verification
• user preference understanding

Innovation

• verifiable reasoning
• generative recommendation
• large language models
• reason-verify-recommend
• multi-dimensional verification