English

Token-Weighted Multi-Target Learning for Generative Recommenders with Curriculum Learning

Information Retrieval 2026-01-27 v1

Abstract

Generative recommender systems have recently attracted attention by formulating next-item prediction as an autoregressive sequence generation task. However, most existing methods optimize standard next-token likelihood and implicitly treat all tokens as equally informative, which is misaligned with semantic-ID-based generation. Accordingly, we propose two complementary information-gain-based token-weighting strategies tailored to generative recommendation with semantic IDs. Front-Greater Weighting captures conditional semantic information gain by prioritizing early tokens that most effectively reduce candidate-item uncertainty given their prefixes and encode coarse semantics. Frequency Weighting models marginal information gain under long-tailed item and token distributions, upweighting rare tokens to counteract popularity bias. Beyond individual strategies, we introduce a multi-target learning framework with curriculum learning that jointly optimizes the two token-weighted objectives alongside standard likelihood, enabling stable optimization and adaptive emphasis across training stages. Extensive experiments on benchmark datasets show that our method consistently outperforms strong baselines and existing token-weighting approaches, with improved robustness, strong generalization across different semantic-ID constructions, and substantial gains on both head and tail items. Code is available at https://github.com/CHIUWEINING/Token-Weighted-Multi-Target-Learning-for-Generative-Recommenders-with-Curriculum-Learning.

Keywords

Cite

@article{arxiv.2601.17787,
  title  = {Token-Weighted Multi-Target Learning for Generative Recommenders with Curriculum Learning},
  author = {Wei-Ning Chiu and Chuan-Ju Wang and Pu-Jen Cheng},
  journal= {arXiv preprint arXiv:2601.17787},
  year   = {2026}
}

Comments

11 pages, 5 figures

R2 v1 2026-07-01T09:19:06.441Z