๐ŸŽ“How I Study AIHISA
๐Ÿ“–Read
๐Ÿ“„Papers๐Ÿ“ฐBlogs๐ŸŽฌCourses
๐Ÿ’กLearn
๐Ÿ›ค๏ธPaths๐Ÿ“šTopics๐Ÿ’กConcepts๐ŸŽดShorts
๐ŸŽฏPractice
๐ŸงฉProblems๐ŸŽฏPrompts๐Ÿง Review
Search
How I Study AI - Learn AI Papers & Lectures the Easy Way

Papers1

AllBeginnerIntermediateAdvanced
All SourcesarXiv
#Masked Autoencoding

Rethinking Generative Recommender Tokenizer: Recsys-Native Encoding and Semantic Quantization Beyond LLMs

Intermediate
Yu Liang, Zhongjin Zhang et al.Feb 2arXiv

This paper proposes ReSID, a new way to turn items into short token codes (Semantic IDs) that are much easier for a recommender to predict.

#Semantic IDs#Generative Recommendation#Representation Learning