Decoupled Temporal Encoding for Generative Recommendation

• URL: http://arxiv.org/abs/2608.16274v1
• Date: Mon, 17 Aug 2026 08:47:21 GMT
• Title: Decoupled Temporal Encoding for Generative Recommendation
• Abstract summary: Positional encoding is a fundamental component of Transformer-based generative recommendation models.<n>Our work is motivated by a real-world food delivery and instant retail recommendation system.<n>We propose Decoupled Temporal, a lightweight framework for generative recommendation.
• Score: 6.5465081246859915
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Positional encoding is a fundamental component of Transformer-based generative recommendation models, where user histories are modeled as autoregressive item sequences. Most positional encoding methods are inherited from natural language processing and mainly represent discrete item order. However, recommendation sequences go beyond ordered lists, as timestamps and temporal effects also shape item relations. Our work is motivated by a real-world food delivery and instant retail recommendation system, where user behavior exhibits multi-level temporal regularities, including recency effects, meal-time peaks, weekday-weekend shifts, and promotion-driven traffic bursts. Existing methods partially address this issue through timestamp features, interval embeddings, decay functions, or attention biases, but they usually inject heterogeneous temporal signals through a unified representation or a single modeling pathway, making it difficult to distinguish broad temporal dynamics from local order cues. To address this limitation, we propose Decoupled Temporal Encoding, a lightweight framework for generative recommendation. DTE separates temporal dynamics from order information through two complementary modules: a personalized macro-temporal module that injects compact temporal primitives into item embeddings, and a time-gated micro-sequential module that introduces relative-order bias only when interactions are temporally dense. DTE is also parameter-efficient and deployment-friendly, allowing easy integration into existing systems.

Related papers

• Beyond Item Order: Temporal Gap Tokenization for Generative Recommendation with Semantic IDs [22.518121739900284]
  ChronoSID is a lightweight temporal augmentation framework for semantic generative recommendation.<n>We discretize historical interaction intervals into fixed log-scale gap tokens and interleave them with semantic ID tokens.<n> Experiments on Amazon benchmarks show that ChronoSID consistently improves over ReSID and other competitive generative recommendation baselines.
  arXiv  Detail & Related papers  (2026-07-04T15:22:30Z)
• POEM: Partial-Order Enhanced Real-Time Sequential Modeling for Recommendation [47.301549241806185]
  We propose a new real-time sequential modeling framework built upon intrinsic partial-order relations from the recommendation cascade.<n>POEM takes real-time multi-task ranking scores generated by upstream ranking modules as supervision to construct dynamic partial-order sequences.<n>Fully deployed on Kuaishou online traffic, POEM achieves significant online gains.
  arXiv  Detail & Related papers  (2026-06-29T08:23:04Z)
• ChronoID: Infusing Explicit Temporal Signals into Semantic IDs for Generative Recommendation [44.412346636470325]
  We investigate where and how the explicit time should be incorporated into semantic ID for generative recommendation.<n>We present a unified framework, ChronoID, for time-aware semantic ID learning.<n>By contributing a new time-explicit generation recommendation benchmark, ChronoID answers the questions: what is the effective way of time, how to design the architecture, and where does the gain come from.
  arXiv  Detail & Related papers  (2026-06-12T08:45:04Z)
• TIE: Time Interval Encoding for Video Generation over Events [50.66585165263848]
  Director-style prompting, robotic action prediction, and interactive video agents demand temporal grounding over concurrent events.<n>Modern video generators represent time as discrete points through point-wise positional encodings.<n>Time Interval TIE is a principled, plug-and-play interval-aware generalization.
  arXiv  Detail & Related papers  (2026-05-11T13:23:14Z)
• RoTE: Coarse-to-Fine Multi-Level Rotary Time Embedding for Sequential Recommendation [17.10655350843282]
  We propose RoTE, a novel multi-level temporal embedding module that explicitly models time span information in sequential recommendation.<n>RoTE decomposes each interaction timestamp into multiple temporal granularities, ranging from coarse to fine, and incorporates the resulting temporal representations into item embeddings.
  arXiv  Detail & Related papers  (2026-04-15T01:38:41Z)
• Discrete Prototypical Memories for Federated Time Series Foundation Models [62.22804699052539]
  Leveraging Large Language Models as federated learning (FL)-based time series foundation models offers a promising way to transfer the generalization capabilities of LLMs to time series data.<n>However, the semantic misalignment between time-series data and the text-centric latent space of existing LLMs often leads to degraded performance.<n>We propose textscFeDPM, a federated framework for time-series foundation models based on discrete prototypical memories.
  arXiv  Detail & Related papers  (2026-04-06T06:57:08Z)
• MEMTS: Internalizing Domain Knowledge via Parameterized Memory for Retrieval-Free Domain Adaptation of Time Series Foundation Models [51.506429027626005]
  Memory for Time Series (MEMTS) is a lightweight and plug-and-play method for retrieval-free domain adaptation in time series forecasting.<n>Key component of MEMTS is a Knowledge Persistence Module (KPM), which internalizes domain-specific temporal dynamics.<n>This paradigm shift enables MEMTS to achieve accurate domain adaptation with constant-time inference and near-zero latency.
  arXiv  Detail & Related papers  (2026-02-14T14:00:06Z)
• UniDiff: A Unified Diffusion Framework for Multimodal Time Series Forecasting [90.47915032778366]
  We propose UniDiff, a unified diffusion framework for multimodal time series forecasting.<n>At its core lies a unified and parallel fusion module, where a single cross-attention mechanism integrates structural information from timestamps and semantic context from texts.<n>Experiments on real-world benchmark datasets across eight domains demonstrate that the proposed UniDiff model achieves state-of-the-art performance.
  arXiv  Detail & Related papers  (2025-12-08T05:36:14Z)
• EntroPE: Entropy-Guided Dynamic Patch Encoder for Time Series Forecasting [50.794700596484894]
  We propose EntroPE (Entropy-Guided Dynamic Patch), a novel, temporally informed framework that dynamically detects transition points via conditional entropy.<n>This preserves temporal structure while retaining the computational benefits of patching.<n> Experiments across long-term forecasting benchmarks demonstrate that EntroPE improves both accuracy and efficiency.
  arXiv  Detail & Related papers  (2025-09-30T12:09:56Z)
• Learning Sequence Representations by Non-local Recurrent Neural Memory [61.65105481899744]
  We propose a Non-local Recurrent Neural Memory (NRNM) for supervised sequence representation learning. Our model is able to capture long-range dependencies and latent high-level features can be distilled by our model. Our model compares favorably against other state-of-the-art methods specifically designed for each of these sequence applications.
  arXiv  Detail & Related papers  (2022-07-20T07:26:15Z)
• Continuous-Time Sequential Recommendation with Temporal Graph Collaborative Transformer [69.0621959845251]
  We propose a new framework Temporal Graph Sequential Recommender (TGSRec) upon our defined continuous-time bi-partite graph. TCT layer can simultaneously capture collaborative signals from both users and items, as well as considering temporal dynamics inside sequential patterns. Empirical results on five datasets show that TGSRec significantly outperforms other baselines.
  arXiv  Detail & Related papers  (2021-08-14T22:50:53Z)
• TEA: Temporal Excitation and Aggregation for Action Recognition [31.076707274791957]
  We propose a Temporal Excitation and Aggregation block, including a motion excitation module and a multiple temporal aggregation module. For short-range motion modeling, the ME module calculates the feature-level temporal differences fromtemporal features. The MTA module proposes to deform the local convolution to a group of sub-convolutions, forming a hierarchical residual architecture.
  arXiv  Detail & Related papers  (2020-04-03T06:53:30Z)

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.