REVIEW 2 cited by
Cold-Starts in Generative Recommendation: A Reproducibility Study
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Cold-Starts in Generative Recommendation: A Reproducibility Study
read the original abstract
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.
Forward citations
Cited by 2 Pith papers
-
VaLiDRec: Variable-Length LLM-Aligned Semantic IDs for Generative Recommendation
VaLiDRec constructs variable-length item IDs from native LLM vocabulary tokens and predicts them in parallel, outperforming fixed-code semantic-ID recommenders on four Amazon datasets.
-
Can Generative Recommendation Reach Cold Items? A Temporal Perspective on Semantic-ID Generation
Under a strict time-ordered evaluation, semantic-ID generative recommenders nearly always fail to retrieve items that appear after training, succeeding only for cold items with already-observed tokens and deep prefixes.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.