works on
feature interaction 1industrial recommender deployment 1large-scale recommendation 1sample-level tokenization 1transformer models 1

From the 1 of 1 paper with an AI index.

researchers with a paper here
collaborating institutions

1 paper

cs.IR2026

Sample Is Feature: Beyond Item-Level, Toward Sample-Level Tokens for Unified Large Recommender Models

Shuli Wang, Junwei Yin, Changhao Li +6

The paper introduces SIF, a method that converts each historical user interaction sample into a token using hierarchical group-adaptive quantization and then mixes these tokens wit…