CapsID: Soft-Routed Variable-Length Semantic IDs for Generative Recommendation

• URL: http://arxiv.org/abs/2605.05096v1
• Date: Wed, 06 May 2026 16:33:13 GMT
• Title: CapsID: Soft-Routed Variable-Length Semantic IDs for Generative Recommendation
• Abstract summary: Generative recommendation maps each item to a sequence of Semantic IDs (SIDs) and recasts retrieval as autoregressive token generation.<n> CAPSID replaces hard residual quantization with capsule routing.<n>On Amazon Beauty, Sports, Toys, and a 35M-item proprietary industrial catalog, CAPSID+SEMANTICBPE improves Recall at 10 by 9.6% on average over ReSID.
• Score: 6.439016735163573
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Generative recommendation maps each item to a sequence of Semantic IDs (SIDs) and recasts retrieval as autoregressive token generation. In this paradigm the main bottleneck is the tokenizer rather than the Transformer: residual vector quantization with a hard nearest-neighbor assignment at every layer collapses multi-faceted item semantics at cluster boundaries and propagates early errors to later SID positions. A common workaround is to append a dense vector or attribute prefix to the SID, but this dual-representation design inflates inference cost and gives up the simplicity of a generative interface. We address the bottleneck at the tokenizer itself. CAPSID replaces hard residual quantization with capsule routing: at each layer an item probabilistically routes to several semantic capsules, the residual is updated by the routed reconstruction rather than by a single winning code, and the SID terminates once the active capsule's confidence is high enough. On top of CAPSID, SEMANTICBPE composes adjacent SID tokens into reusable subwords by combining their co-occurrence with their embedding compatibility. On Amazon Beauty, Sports, Toys, and a 35M-item proprietary industrial catalog, CAPSID+SEMANTICBPE improves Recall at 10 by 9.6% on average over ReSID, the strongest single-representation baseline, and matches or exceeds a COBRA-style sparse-dense system on every public benchmark while running at 51% of its inference latency. Ablations show that soft routing, iterative agreement, and confidence-driven length each contribute independently, and the gains are largest on tail items where boundary semantics dominate.

Related papers

• Preference Shapes Relevance: Cross-component Hierarchical Semantic Alignment for Personalized Generative Retrieval [37.821022474615994]
  Generative Retrieval (GR) has emerged as a promising paradigm by mapping queries directly to Semantic IDs (SIDs) with powerful representation capabilities for candidate items.<n>Existing SIDs derived solely from item content create a semantic gap, failing to align dynamic query intents with static item representations.<n>We present a novel personalized GR framework that models user behavior by synergizing discrete SIDs for structural guidance and continuous representations for fine-grained semantic refinement.
  arXiv  Detail & Related papers  (2026-08-31T10:30:22Z)
• Rethinking Item Tokenization in Generative Recommenders: From Fixed Atoms to Semantic Subwords [50.621722853036005]
  In generative recommender systems, items are typically tokenized into fixed-length semantic ID sequences for autoregressive next-item prediction.<n>For user-context modeling, this fine-grained representation triggers Intra-item Attention Overload.<n>We propose Semantic Subword Tokenization (SST), which represents historical items as variable-length semantic subwords.
  arXiv  Detail & Related papers  (2026-08-24T02:42:51Z)
• Hierarchical Residual Policy Optimization for Generative Recommendations [46.030303420508254]
  We propose a post-training framework that converts item-level outcomes into dense, token-aligned learning signals.<n>Experiments on a public dataset and an online A/B test in a large-scale commercial system show consistent gains in session-level utility and key business metrics.
  arXiv  Detail & Related papers  (2026-08-01T16:33:15Z)
• Gryphon: A Unified Architecture for Semantic-ID Generation and Item-Level Scoring in Industrial Recommendations [0.0]
  We introduce a jointly trained item-level scoring component alongside SID generation.<n>Instead of ranking SIDs by accumulated token likelihood, Gryphon resolves each generated SID to its concrete items.<n>On an industrial music service, with item-level scoring trained under a next-item-prediction objective, Gryphon attains the highest item-level Recall@1000.
  arXiv  Detail & Related papers  (2026-06-07T12:31:26Z)
• DREAM: Dynamic Refinement of Early Assignment Mappings [7.781421530363403]
  Current methods assign each item a single static identifier through offline tokenization before sufficient user feedback is observed.<n>We identify this one-shot commitment as the fundamental cold-start bottleneck in SID-based generative recommendation.<n>We propose DREAM, a three-stage framework that resolves this flaw through progressive refinement.
  arXiv  Detail & Related papers  (2026-06-05T06:21:01Z)
• Conditional Memory Enhanced Item Representation for Generative Recommendation [46.59623438612085]
  Generative recommendation (GR) has emerged as a promising paradigm that predicts target items by autoregressively generating their semantic identifiers (SID)<n>Most GR methods follow a quantization-representation-generation pipeline, first assigning each item a SID, then constructing input representations from SID-token embeddings, and finally predicting the target SID through autoregressive generation.<n>We propose ComeIR, a Conditional Memory enhanced Item Representation framework that reconstructs SID-token embeddings into item-aware inputs and restores the token granularity during SID decoding.
  arXiv  Detail & Related papers  (2026-05-12T02:56:59Z)
• UniRec: Bridging the Expressive Gap between Generative and Discriminative Recommendation via Chain-of-Attribute [12.89120699793625]
  Generative Recommendation (GR) reframes retrieval and ranking as autoregressive decoding over Semantic IDs (SIDs)<n>We formalize this via Bayes' theorem: ranking by p(y|f,u) is equivalent to ranking by p(f|y,u), which factorizes autoregressively over item features.<n>We propose UniRec with Chain-of-Attribute (CoA) as its core mechanism. CoA prefixes each SID sequence with structured attribute tokens:category, seller, brand, before decoding the SID, recovering the item-side feature crossing that discriminative models exploit.
  arXiv  Detail & Related papers  (2026-04-14T03:13:50Z)
• One Model, Many Budgets: Elastic Latent Interfaces for Diffusion Transformers [80.19461768457622]
  Elastic Latent Interface Transformer (ELIT) is a drop-in, DiT-compatible mechanism that decouples input image size from compute.<n>Read and Write cross-attention layers move information between spatial tokens and latents and prioritize important input regions.<n>On ImageNet-1K 512px, ELIT delivers an average gain of $35.3%$ and $39.6%$ in FID and FDD scores.
  arXiv  Detail & Related papers  (2026-03-12T17:57:04Z)
• End-to-End Semantic ID Generation for Generative Advertisement Recommendation [33.453121305193434]
  We propose a Unified SID generation framework for generative advertisement recommendation.<n>Specifically, we jointly optimize embeddings and SIDs in an end-to-end manner from raw advertising data.<n>Experiments demonstrate that UniSID consistently outperforms state-of-the-art SID generation methods.
  arXiv  Detail & Related papers  (2026-02-11T02:38:26Z)
• GLASS: A Generative Recommender for Long-sequence Modeling via SID-Tier and Semantic Search [51.44490997013772]
  GLASS is a novel framework that integrates long-term user interests into the generative process via SID-Tier and Semantic Search.<n>We show that GLASS outperforms state-of-the-art baselines in experiments on two large-scale real-world datasets.
  arXiv  Detail & Related papers  (2026-02-05T13:48:33Z)
• Differentiable Semantic ID for Generative Recommendation [65.83703273297492]
  Generative recommendation provides a novel paradigm in which each item is represented by a discrete semantic ID (SID) learned from rich content.<n>In practice, SIDs are typically optimized only for content reconstruction rather than recommendation accuracy.<n>A natural approach is to make semantic indexing differentiable so that recommendation gradients can directly influence SID learning.<n>We propose DIGER, a first step toward effective differentiable semantic IDs for generative recommendation.
  arXiv  Detail & Related papers  (2026-01-27T15:34:11Z)
• DiffGRM: Diffusion-based Generative Recommendation Model [63.35379395455103]
  Generative recommendation (GR) is an emerging paradigm that represents each item via a tokenizer as an n-digit semantic ID (SID)<n>We propose DiffGRM, a diffusion-based GR model that replaces the autoregressive decoder with a masked discrete diffusion model (MDM)<n> Experiments show consistent gains over strong generative and discriminative recommendation baselines on multiple datasets.
  arXiv  Detail & Related papers  (2025-10-21T03:23:32Z)
• HAT: Hierarchical Aggregation Transformers for Person Re-identification [87.02828084991062]
  We take advantages of both CNNs and Transformers for image-based person Re-ID with high performance. Work is the first to take advantages of both CNNs and Transformers for image-based person Re-ID.
  arXiv  Detail & Related papers  (2021-07-13T09:34:54Z)

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.