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

2026-04-16

AI Summary

Existing recommender systems struggle to fully exploit sample-level information, and the structural heterogeneity between sequential and non-sequential features limits model expressiveness. To address this, this work proposes a “sample-as-feature” paradigm that directly encodes raw samples into sample-level tokens within sequences. A hierarchical group-adaptive quantization (HGAQ) mechanism is introduced to construct a sample tokenizer, and a novel SIF-Mixer module is designed to jointly model token-level and sample-level interactions. This approach enables full-sample contextual modeling and deep homogeneous feature interaction while preserving time-varying sample-level attributes—a capability not previously achieved. Evaluated on Meituan Waimai’s large-scale dataset, the method significantly outperforms existing approaches in recommendation performance and has been successfully deployed in production.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Scaling industrial recommender models has followed two parallel paradigms: \textbf{sample information scaling} -- enriching the information content of each training sample through deeper and longer behavior sequences -- and \textbf{model capacity scaling} -- unifying sequence modeling and feature interaction within a single Transformer backbone. However, these two paradigms still face two structural limitations. Firstly, sample information scaling methods encode only a subset of each historical interaction into the sequence token, leaving the majority of the original sample context unexploited and precluding the modeling of sample-level, time-varying features. Secondly, model capacity scaling methods are inherently constrained by the structural heterogeneity between sequential and non-sequential features, preventing the model from fully realizing its representational capacity. To address these issues, we propose \textbf{SIF} (\emph{Sample Is Feature}), which encodes each historical Raw Sample directly into the sequence token -- maximally preserving sample information while simultaneously resolving the heterogeneity between sequential and non-sequential features. SIF consists of two key components. The \textbf{Sample Tokenizer} quantizes each historical Raw Sample into a Token Sample via hierarchical group-adaptive quantization (HGAQ), enabling full sample-level context to be incorporated into the sequence efficiently. The \textbf{SIF-Mixer} then performs deep feature interaction over the homogeneous sample representations via token-level and sample-level mixing, fully unleashing the model's representational capacity. Extensive experiments on a large-scale industrial dataset validate SIF's effectiveness, and we have successfully deployed SIF on the Meituan food delivery platform.

Problem

• sample-level tokens
• feature heterogeneity
• sequence modeling
• recommendation systems
• model scaling

Innovation

• Sample Is Feature
• Sample Tokenizer
• SIF-Mixer
• hierarchical group-adaptive quantization
• unified recommender model