TokenFormer unifies multi-field and sequential recommendation modeling via bottom-full-top-sliding attention and non-linear interaction representations to avoid sequential collapse and deliver state-of-the-art performance.
Title resolution pending
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 6roles
background 1polarities
background 1representative citing papers
UG-Separation framework disentangles user-side and item-side flows in TokenMixer dense-interaction models to enable reusable user computations, cutting inference latency up to 20% in ByteDance production scenarios.
MBGR is a new generative recommendation framework using business-aware semantic IDs, multi-business prediction, and label dynamic routing to handle multiple businesses without seesaw effects or representation confusion, validated by experiments and deployed at Meituan.
OCARM uses teacher-student distillation to let retention models learn from inaccessible post-conversion content without feature leakage, yielding improvements in offline experiments and online A/B tests.
A2Gen treats timed user actions on short videos as generative sequences and reports large-scale online gains in watch time, interactions, and retention.
citing papers explorer
-
TokenFormer: Unify the Multi-Field and Sequential Recommendation Worlds
TokenFormer unifies multi-field and sequential recommendation modeling via bottom-full-top-sliding attention and non-linear interaction representations to avoid sequential collapse and deliver state-of-the-art performance.
-
Compute Only Once: UG-Separation for Efficient Large Recommendation Models
UG-Separation framework disentangles user-side and item-side flows in TokenMixer dense-interaction models to enable reusable user computations, cutting inference latency up to 20% in ByteDance production scenarios.
-
MBGR: Multi-Business Prediction for Generative Recommendation at Meituan
MBGR is a new generative recommendation framework using business-aware semantic IDs, multi-business prediction, and label dynamic routing to handle multiple businesses without seesaw effects or representation confusion, validated by experiments and deployed at Meituan.
-
Break the Inaccessible Boundary: Distilling Post-Conversion Content for User Retention Modeling
OCARM uses teacher-student distillation to let retention models learn from inaccessible post-conversion content without feature leakage, yielding improvements in offline experiments and online A/B tests.
-
Action-Aware Generative Sequence Modeling for Short Video Recommendation
A2Gen treats timed user actions on short videos as generative sequences and reports large-scale online gains in watch time, interactions, and retention.
- UniScale: Synergistic Entire Space Data and Model Scaling for Search Ranking