aiqbench analysis
This research introduces FOUNDv2, a novel framework that quantizes user representations into discrete tokens to solve storage and integration challenges in large-scale recommendation systems. By leveraging multi-view RQ-VAEs and shared codebooks, the method achieves significant efficiency gains while maintaining predictive accuracy across diverse industrial benchmarks like Alipay's platform. The work advances machine learning infrastructure by demonstrating how tokenization can unify heterogeneous data sources without sacrificing modeling granularity.
FOUNDv2: Learning Unified User Quantized Tokenizers for User Representation
arXiv:2508.00956v3 Announce Type: replace Abstract: User representation learning serves as a fundamental pillar for personalized services on large-scale web platforms. Despite its importance, conventional continuous embedding methods face significant challenges, including the lack of a unified paradigm for multi-source data integration, prohibitive storage overhead due to low information density, and the lack of multi-scale modeling granularity. To overcome these limitations, we introduce FOUNDv2, a comprehensive user representation scheme centered on the Unified User Quantized Tokenizer U2QT) framework. FOUNDv2 transforms heterogeneous user data into a standardized discrete token space through a robust two-stage architecture. Specifically, the framework first extracts compact feature representations and subsequently employs a multi-view RQ-VAE to discretize them into storage-efficient tokens using shared and source-specific codebooks. To empower these representations with predictive intelligence, we further design multi-scale alignment objectives to capture both fine-grained behavioral dependencies and macro-temporal periodicity. Extensive experiments on various benchmarks demonstrate that FOUNDv2 consistently outperforms task-specific baselines while achieving substantial reductions in storage and computational costs. Finally, the large-scale deployment of FOUNDv2 on Alipay validates its practical scalability and efficiency across diverse industrial scenarios. The main code is available at: https://github.com/chuanhe1999/FOUNDv2.
Abstract:User representation learning serves as a fundamental pillar for personalized services on large-scale web platforms. Despite its importance, conventional continuous embedding methods face significant challenges, including the lack of a unified paradigm for multi-source data integration, prohibitive storage overhead due to low information density, and the lack of multi-scale modeling granularity. To overcome these limitations, we introduce FOUNDv2, a comprehensive user representation scheme centered on the Unified User Quantized Tokenizer U2QT) framework. FOUNDv2 transforms heterogeneous user data into a standardized discrete token space through a robust two-stage architecture. Specifically, the framework first extracts compact feature representations and subsequently employs a multi-view RQ-VAE to discretize them into storage-efficient tokens using shared and source-specific codebooks. To empower these representations with predictive intelligence, we further design multi-scale alignment objectives to capture both fine-grained behavioral dependencies and macro-temporal periodicity. Extensive experiments on various benchmarks demonstrate that FOUNDv2 consistently outperforms task-specific baselines while achieving substantial reductions in storage and computational costs. Finally, the large-scale deployment of FOUNDv2 on Alipay validates its practical scalability and efficiency across diverse industrial scenarios. The main code is available at: this https URL.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
arXiv:2508.00956 [cs.LG]
(or arXiv:2508.00956v3 [cs.LG] for this version)
Full text under the publisher's open licence · 200 words
Related coverage
[AINews] Opus 5.5 is good at explainer videos
Latent Space · Sep 29, 2026
[AINews] The Future of Latent Space
Latent Space · Sep 25, 2026
Exploring Hierarchical Interest Representation For Meta Ads Deep Funnel Optimization
Meta Engineering · Jul 15, 2026
EntangleCodec: A Unified Discrete Audio Tokenizer via Semantic-Acoustic Entanglement
arXiv · Jun 3, 2026