Uncertainty-Aware Semantic Decoding for LLM-Based Sequential Recommendation

TL;DR

This paper introduces an Uncertainty-aware Semantic Decoding framework that enhances large language model-based sequential recommendation by clustering items and controlling scoring based on uncertainty, leading to significant performance improvements.

Contribution

It proposes a novel decoding method combining semantic clustering and entropy-based uncertainty control to improve recommendation accuracy in LLM-based systems.

Findings

Achieved 18.5% improvement in HR@3 on Amazon datasets.

Improved NDCG@3 by 11.9% and MRR@3 by 10.8%.

Framework adapts well across different recommendation domains.

Abstract

Large language models have been widely applied to sequential recommendation tasks, yet during inference, they continue to rely on decoding strategies developed for natural language processing. This creates a mismatch between text-generation objectives and recommendation next item selection objectives. This paper addresses this limitation by proposing an Uncertainty-aware Semantic Decoding (USD) framework that combines logit-based clustering with adaptive scoring to improve next-item predictions. Our approach clusters items with similar logit vectors into semantic equivalence groups, then redistributes probability mass within these clusters and computes entropy across them to control item scoring and sampling temperature during recommendation inference. Experiments on Amazon Product datasets (six domains) gains of 18.5\% in HR@3, 11.9\% in NDCG@3, and 10.8\% in MRR@3 compared to state-of-the-art baselines. Hyperparameter analysis confirms the optimal parameters among various settings, and experiments on H\&M, and Netflix datasets indicate that the framework can adapt to differing recommendation domains. The experimental results confirm that integrating semantic clustering and uncertainty assessment yields more reliable and accurate recommendations.