# Decoding Matters: Addressing Amplification Bias and Homogeneity Issue for LLM-based Recommendation

arXiv: https://arxiv.org/abs/2406.14900 · HTML: https://mlsanity.com/paper/2406.14900/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Jun 2024.

## Built on by (most cited first, 11 of 11)

- [Reinforced Latent Reasoning for LLM-based Recommendation](https://mlsanity.com/paper/2505.19092/) (2025)
- [MiniOneRec: An Open-Source Framework for Scaling Generative Recommendation](https://mlsanity.com/paper/2510.24431/) (2025)
- [Toward Holistic Evaluation of Recommender Systems Powered by Generative Models](https://mlsanity.com/paper/2504.06667/) (2025)
- [Causality-Enhanced Behavior Sequence Modeling in LLMs for Personalized Recommendation](https://mlsanity.com/paper/2410.22809/) (2024)
- [Societal Impacts Research Requires Benchmarks for Creative Composition Tasks](https://mlsanity.com/paper/2504.06549/) (2025)
- [Spend Search Where It Pays: Value-Guided Structured Sampling and Optimization for Generative Recommendation](https://mlsanity.com/paper/2602.10699/) (2026)
- [Verifiable Reasoning for LLM-based Generative Recommendation](https://mlsanity.com/paper/2603.07725/) (2026)
- [Prompt Optimization Across Multiple Agents for Representing Diverse Human Populations](https://mlsanity.com/paper/2510.07064/) (2025)
- [Reasoning to Rank: An End-to-End Solution for Exploiting Large Language Models for Recommendation](https://mlsanity.com/paper/2602.12530/) (2026)
- [Reinforced Preference Optimization for Reasoning-Augmented Recommendations](https://mlsanity.com/paper/2605.21967/) (2026)
- [Difficulty-Aware Semantic-ID Optimization for Generative Recommendation](https://mlsanity.com/paper/2608.20611/) (2026)

## Abstract (arXiv)

Adapting Large Language Models (LLMs) for recommendation requires careful consideration of the decoding process, given the inherent differences between generating items and natural language. Existing approaches often directly apply LLMs' original decoding methods. However, we find these methods encounter significant challenges: 1) amplification bias -- where standard length normalization inflates scores for items containing tokens with generation probabilities close to 1 (termed ghost tokens), and 2) homogeneity issue -- generating multiple similar or repetitive items for a user. To tackle these challenges, we introduce a new decoding approach named Debiasing-Diversifying Decoding (D3). D3 disables length normalization for ghost tokens to alleviate amplification bias, and it incorporates a text-free assistant model to encourage tokens less frequently generated by LLMs for counteracting recommendation homogeneity. Extensive experiments on real-world datasets demonstrate the method's effectiveness in enhancing accuracy and diversity. The code is available at https://github.com/SAI990323/DecodingMatters.