KUAISHOU Explorer LLM-Rec Challenge 2026: Reasoning Generative Recommendation

2026-09-30

AI Summary

This study addresses the limited performance gains of chain-of-thought reasoning in generative recommendation by proposing the OneReason model. Building upon the OneRec unified representation space, the approach enhances semantic alignment between item identifiers and natural language through structured template supervision and reinforcement learning, thereby unlocking latent reasoning capabilities. By integrating large language models, autoregressive prediction, and advanced reinforcement learning algorithms, the proposed method substantially improves the performance of reasoning-based generative recommendation. Beyond validating its effectiveness, this project co-organized a challenge with the SIGIR community to promote frontier research in this emerging domain.

Abstract

Generative recommendation, has been attracted a surge of attentions in industrial and academic research community, towards to build more smart system to build next-generation recommender. Under the significant developing wave of large language model, our team have been developed Semantic ID based OneRec/OneRec-V2. These models have been widely deployed in production and demonstrate the scaling potential of the autoregressive next-item prediction paradigm for industrial recommender systems. Building on the success of OneRec, we further explored a series of models, including OneRec-Think, OpenOneRec, and OneReason, that connect item Semantic IDs with natural language in a unified representation space and seek to unlock the potential of natural-language chain-of-thought (CoT) reasoning for recommendation. However, our preliminary works found that introducing reasoning CoT does not always improve the recommendation performance. To address this issue, OneReason strengthens the semantic alignment between items and language, introduces structured template-based supervision for interest reasoning, and applies advanced reinforcement learning techniques to make reasoning more beneficial to recommendation. As a frontier topic to building recommendation foundation models, we believe this topic has significant research value and hope to encourage more researchers to explore it together. To this end, together with the SIGIR 2026 community, we organized the KUAISHOU Explorer LLM-Rec Challenge 2026: Reasoning Generative Recommendation.

Problem

• Generative Recommendation
• Chain-of-Thought Reasoning
• Large Language Model
• Semantic ID

Innovation

• Generative Recommendation
• Semantic ID
• Chain-of-Thought Reasoning
• Reinforcement Learning
• Large Language Model