Generative recommender systems using COSETTE tokenization and the MARIUS architecture reach or exceed the accuracy of a strong ID-based SASRec baseline on standard Amazon benchmarks.
Recommendation as language processing (RLP): A unified pretrain, personalized prompt & predict paradigm (P5)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.IR 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Closing the Performance Gap in Generative Recommenders with Collaborative Tokenization and Efficient Modeling
Generative recommender systems using COSETTE tokenization and the MARIUS architecture reach or exceed the accuracy of a strong ID-based SASRec baseline on standard Amazon benchmarks.