Scaling Generative Recommendations with Context Parallelism on Hierarchical Sequential Transducers

• URL: http://arxiv.org/abs/2508.04711v2
• Date: Sat, 16 Aug 2025 00:20:02 GMT
• Title: Scaling Generative Recommendations with Context Parallelism on Hierarchical Sequential Transducers
• Authors: Yue Dong, Han Li, Shen Li, Nikhil Patel, Xing Liu, Xiaodong Wang, Chuanhao Zhuge,
• Abstract summary: We introduce context parallelism with jagged tensor support for HSTU attention, establishing foundational capabilities for scaling up sequence dimensions.<n>Our approach enables a 5.3x increase in supported user interaction sequence length, while achieving a 1.55x scaling factor when combined with Distributed Data Parallelism (DDP)
• Score: 29.05624030090006
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Large-scale recommendation systems are pivotal to process an immense volume of daily user interactions, requiring the effective modeling of high cardinality and heterogeneous features to ensure accurate predictions. In prior work, we introduced Hierarchical Sequential Transducers (HSTU), an attention-based architecture for modeling high cardinality, non-stationary streaming recommendation data, providing good scaling law in the generative recommender framework (GR). Recent studies and experiments demonstrate that attending to longer user history sequences yields significant metric improvements. However, scaling sequence length is activation-heavy, necessitating parallelism solutions to effectively shard activation memory. In transformer-based LLMs, context parallelism (CP) is a commonly used technique that distributes computation along the sequence-length dimension across multiple GPUs, effectively reducing memory usage from attention activations. In contrast, production ranking models typically utilize jagged input tensors to represent user interaction features, introducing unique CP implementation challenges. In this work, we introduce context parallelism with jagged tensor support for HSTU attention, establishing foundational capabilities for scaling up sequence dimensions. Our approach enables a 5.3x increase in supported user interaction sequence length, while achieving a 1.55x scaling factor when combined with Distributed Data Parallelism (DDP).

Related papers

• Beyond the Flat Sequence: Hierarchical and Preference-Aware Generative Recommendations [35.58864660038236]
  We propose a novel framework named HPGR (Hierarchical and Preference-aware Generative Recommender)<n>First, a structure-aware pre-training stage employs a session-based Masked Item Modeling objective to learn a hierarchically-informed and semantically rich item representation space.<n>Second, a preference-aware fine-tuning stage leverages these powerful representations to implement a Preference-Guided Sparse Attention mechanism.
  arXiv  Detail & Related papers  (2026-03-01T08:15:34Z)
• MixFormer: Co-Scaling Up Dense and Sequence in Industrial Recommenders [11.566232697512879]
  MixFormer is a unified Transformer-style architecture tailored for recommender systems.<n>It jointly models sequential behaviors and feature interactions within a single backbone.<n>Experiments on large-scale industrial datasets demonstrate that MixFormer consistently exhibits superior accuracy and efficiency.
  arXiv  Detail & Related papers  (2026-02-15T11:53:30Z)
• GEMs: Breaking the Long-Sequence Barrier in Generative Recommendation with a Multi-Stream Decoder [54.64137490632567]
  We propose a novel and unified framework designed to capture users' sequences from long-term history.<n>Generative Multi-streamers ( GEMs) break user sequences into three streams.<n>Extensive experiments on large-scale industrial datasets demonstrate that GEMs significantly outperforms state-the-art methods in recommendation accuracy.
  arXiv  Detail & Related papers  (2026-02-14T06:42:56Z)
• Compress, Cross and Scale: Multi-Level Compression Cross Networks for Efficient Scaling in Recommender Systems [5.897678894426804]
  MLCC is a structured feature interaction architecture that organizes feature crosses through hierarchical compression and dynamic composition.<n>MC-MLCC is a Multi-Channel extension that decomposes feature interactions into parallel subspaces.<n>Our proposed models consistently outperform strong DLRM-style baselines by up to 0.52 AUC, while reducing model parameters and FLOPs by up to 26$times$ under comparable performance.
  arXiv  Detail & Related papers  (2026-02-12T15:06:46Z)
• PRISM: Parallel Residual Iterative Sequence Model [52.26239951489612]
  We propose PRISM (Parallel Residual Iterative Sequence Model) to resolve this tension.<n>PRISM introduces a solver-inspired inductive bias that captures key structural properties of multi-step refinement in a parallelizable form.<n>We prove that this formulation achieves Rank-$L$ accumulation, structurally expanding the update manifold beyond the single-step Rank-$1$ bottleneck.
  arXiv  Detail & Related papers  (2026-02-11T12:39:41Z)
• HyFormer: Revisiting the Roles of Sequence Modeling and Feature Interaction in CTR Prediction [8.97787361529607]
  This paper presents HyFormer, a unified hybrid transformer architecture that tightly integrates long-sequence modeling and feature interaction into a single backbone.<n>Experiments on billion-scale industrial datasets demonstrate that HyFormer consistently outperforms strong LONGER and RankMixer baselines.
  arXiv  Detail & Related papers  (2026-01-19T02:55:05Z)
• Higher-order Linear Attention [59.92962330635185]
  quadratic cost of scaled dot-product attention is a central obstacle to scaling autoregressive language models to long contexts.<n>We introduce Higher-order Linear Attention (HLA), a causal, streaming mechanism that realizes higher interactions via compact prefix sufficient statistics.
  arXiv  Detail & Related papers  (2025-10-31T07:54:37Z)
• Sequential-Parallel Duality in Prefix Scannable Models [68.39855814099997]
  Recent developments have given rise to various models, such as Gated Linear Attention (GLA) and Mamba.<n>This raises a natural question: can we characterize the full class of neural sequence models that support near-constant-time parallel evaluation and linear-time, constant-space sequential inference?
  arXiv  Detail & Related papers  (2025-06-12T17:32:02Z)
• Quantifying Memory Utilization with Effective State-Size [73.52115209375343]
  We develop a measure of textitmemory utilization'<n>This metric is tailored to the fundamental class of systems with textitinput-invariant and textitinput-varying linear operators
  arXiv  Detail & Related papers  (2025-04-28T08:12:30Z)
• A Novel Mamba-based Sequential Recommendation Method [4.941272356564765]
  Sequential recommendation (SR) encodes user activity to predict the next action.<n> Transformer-based models have proven effective for sequential recommendation, but the complexity of the self-attention module in Transformers scales quadratically with the sequence length.<n>We propose a novel multi-head latent Mamba architecture, which employs multiple low-dimensional Mamba layers and fully connected layers.
  arXiv  Detail & Related papers  (2025-04-10T02:43:19Z)
• Tensor Product Attention Is All You Need [53.69820973900921]
  Product Attention (TPA) is a novel attention mechanism that uses tensor decompositions to represent queries, keys, and values compactly.<n>TPA achieves improved model quality alongside memory efficiency.<n>Based on TPA, we introduce the Product Attention Transformer,(T6), a new model architecture for sequence modeling.
  arXiv  Detail & Related papers  (2025-01-11T03:37:10Z)
• Continual Low-Rank Scaled Dot-product Attention [67.11704350478475]
  We introduce a new formulation of the Scaled-product Attention based on the Nystr"om approximation that is suitable for Continual Inference.<n>In experiments on Online Audio Classification and Online Action Detection tasks, the proposed Continual Scaled Dot-product Attention can lower the number of operations by up to three orders of magnitude.
  arXiv  Detail & Related papers  (2024-12-04T11:05:01Z)
• Optimizing Sequential Recommendation Models with Scaling Laws and Approximate Entropy [104.48511402784763]
  Performance Law for SR models aims to theoretically investigate and model the relationship between model performance and data quality.<n>We propose Approximate Entropy (ApEn) to assess data quality, presenting a more nuanced approach compared to traditional data quantity metrics.
  arXiv  Detail & Related papers  (2024-11-30T10:56:30Z)
• ELASTIC: Efficient Linear Attention for Sequential Interest Compression [5.689306819772134]
  State-of-the-art sequential recommendation models heavily rely on transformer's attention mechanism.<n>We propose ELASTIC, an Efficient Linear Attention for SequenTial Interest Compression.<n>We conduct extensive experiments on various public datasets and compare it with several strong sequential recommenders.
  arXiv  Detail & Related papers  (2024-08-18T06:41:46Z)
• Supervised Learning for Non-Sequential Data: A Canonical Polyadic Decomposition Approach [85.12934750565971]
  Efficient modelling of feature interactions underpins supervised learning for non-sequential tasks. To alleviate this issue, it has been proposed to implicitly represent the model parameters as a tensor. For enhanced expressiveness, we generalize the framework to allow feature mapping to arbitrarily high-dimensional feature vectors.
  arXiv  Detail & Related papers  (2020-01-27T22:38:40Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.