Cold-Starts in Generative Recommendation: A Reproducibility Study

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Zhen, Zhao, Jujia, Ma, Xinyu, Xin, Xin, de Rijke, Maarten, Ren, Zhaochun
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913006676344832
author Zhang, Zhen
Zhao, Jujia
Ma, Xinyu
Xin, Xin
de Rijke, Maarten
Ren, Zhaochun
author_facet Zhang, Zhen
Zhao, Jujia
Ma, Xinyu
Xin, Xin
de Rijke, Maarten
Ren, Zhaochun
contents Cold-start recommendation remains a central challenge in dynamic, open-world platforms, requiring models to recommend for newly registered users (user cold-start) and to recommend newly introduced items to existing users (item cold-start) under sparse or missing interaction signals. Recent generative recommenders built on pre-trained language models (PLMs) are often expected to mitigate cold-start by using item semantic information (e.g., titles and descriptions) and test-time conditioning on limited user context. However, cold-start is rarely treated as a primary evaluation setting in existing studies, and reported gains are difficult to interpret because key design choices, such as model scale, identifier design, and training strategy, are frequently changed together. In this work, we present a systematic reproducibility study of generative recommendation under a unified suite of cold-start protocols.
format Preprint
id arxiv_https___arxiv_org_abs_2603_29845
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Cold-Starts in Generative Recommendation: A Reproducibility Study
Zhang, Zhen
Zhao, Jujia
Ma, Xinyu
Xin, Xin
de Rijke, Maarten
Ren, Zhaochun
Information Retrieval
Cold-start recommendation remains a central challenge in dynamic, open-world platforms, requiring models to recommend for newly registered users (user cold-start) and to recommend newly introduced items to existing users (item cold-start) under sparse or missing interaction signals. Recent generative recommenders built on pre-trained language models (PLMs) are often expected to mitigate cold-start by using item semantic information (e.g., titles and descriptions) and test-time conditioning on limited user context. However, cold-start is rarely treated as a primary evaluation setting in existing studies, and reported gains are difficult to interpret because key design choices, such as model scale, identifier design, and training strategy, are frequently changed together. In this work, we present a systematic reproducibility study of generative recommendation under a unified suite of cold-start protocols.
title Cold-Starts in Generative Recommendation: A Reproducibility Study
topic Information Retrieval
url https://arxiv.org/abs/2603.29845