Don't Waste It: Guiding Generative Recommenders with Structured Human Priors via Multi-head Decoding

• URL: http://arxiv.org/abs/2511.10492v2
• Date: Sun, 16 Nov 2025 15:23:07 GMT
• Title: Don't Waste It: Guiding Generative Recommenders with Structured Human Priors via Multi-head Decoding
• Abstract summary: We introduce a backbone-agnostic framework that seamlessly integrates human priors directly into the end-to-end training of generative recommenders.<n>Our method significantly enhances both accuracy and beyond-accuracy objectives.
• Score: 36.10174210922267
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Optimizing recommender systems for objectives beyond accuracy, such as diversity, novelty, and personalization, is crucial for long-term user satisfaction. To this end, industrial practitioners have accumulated vast amounts of structured domain knowledge, which we term human priors (e.g., item taxonomies, temporal patterns). This knowledge is typically applied through post-hoc adjustments during ranking or post-ranking. However, this approach remains decoupled from the core model learning, which is particularly undesirable as the industry shifts to end-to-end generative recommendation foundation models. On the other hand, many methods targeting these beyond-accuracy objectives often require architecture-specific modifications and discard these valuable human priors by learning user intent in a fully unsupervised manner. Instead of discarding the human priors accumulated over years of practice, we introduce a backbone-agnostic framework that seamlessly integrates these human priors directly into the end-to-end training of generative recommenders. With lightweight, prior-conditioned adapter heads inspired by efficient LLM decoding strategies, our approach guides the model to disentangle user intent along human-understandable axes (e.g., interaction types, long- vs. short-term interests). We also introduce a hierarchical composition strategy for modeling complex interactions across different prior types. Extensive experiments on three large-scale datasets demonstrate that our method significantly enhances both accuracy and beyond-accuracy objectives. We also show that human priors allow the backbone model to more effectively leverage longer context lengths and larger model sizes.

Related papers

• Long-term User Engagement Optimization through Model-agnostic Downstream Rewards Learning [2.66348673166416]
  We present a unified, model-agnostic downstream reward framework for optimizing long-term user value in large-scale recommendation systems.<n>We propose several model-agnostic downstream rewards signals derived from observed user action patterns across multiple sources.<n>Online A/B experiments demonstrate consistent improvements in engagement and retention-related metrics.
  arXiv  Detail & Related papers  (2026-07-15T16:17:58Z)
• Unleashing Guidance Without Classifiers for Human-Object Interaction Animation [54.24065561632133]
  Prior diffusion-based approaches often rely on hand-crafted contact priors or human-imposed kinematic constraints to improve contact quality.<n>We propose a data-driven alternative in which guidance emerges from the denoising pace itself, reducing dependence on manually designed priors.<n>We find that this data-driven guidance is inherently contact-aware, and can be enhanced when training is augmented with a broad spectrum of synthetic object geometries.
  arXiv  Detail & Related papers  (2026-03-26T17:59:34Z)
• Synthetic Interaction Data for Scalable Personalization in Large Language Models [67.31884245564086]
  We introduce a high-fidelity synthetic data generation framework called PersonaGym.<n>Unlike prior work that treats personalization as static persona-preference pairs, PersonaGym models a dynamic preference process.<n>We release PersonaAtlas, a large-scale, high-quality, and diverse synthetic dataset of high-fidelity multi-turn personalized interaction trajectories.
  arXiv  Detail & Related papers  (2026-02-12T20:41:22Z)
• Breaking the Likelihood Trap: Consistent Generative Recommendation with Graph-structured Model [8.011282322871958]
  Reranking, as the final stage of recommender systems, demands real-time inference, accuracy, and diversity.<n>We propose Consistent Graph-structured Generative Recommendation (Congrats), a novel generative reranking framework.<n>Congrats has been evaluated on a large-scale video-sharing app, Kuaishou, with over 300 million daily active users.
  arXiv  Detail & Related papers  (2025-10-11T09:21:01Z)
• Slow Thinking for Sequential Recommendation [88.46598279655575]
  We present a novel slow thinking recommendation model, named STREAM-Rec.<n>Our approach is capable of analyzing historical user behavior, generating a multi-step, deliberative reasoning process, and delivering personalized recommendations.<n>In particular, we focus on two key challenges: (1) identifying the suitable reasoning patterns in recommender systems, and (2) exploring how to effectively stimulate the reasoning capabilities of traditional recommenders.
  arXiv  Detail & Related papers  (2025-04-13T15:53:30Z)
• A Survey of Direct Preference Optimization [103.59317151002693]
  Large Language Models (LLMs) have demonstrated unprecedented generative capabilities.<n>Their alignment with human values remains critical for ensuring helpful and harmless deployments.<n>Direct Preference Optimization (DPO) has recently gained prominence as a streamlined alternative.
  arXiv  Detail & Related papers  (2025-03-12T08:45:15Z)
• On-the-fly Preference Alignment via Principle-Guided Decoding [27.50204023448716]
  We introduce On-the-fly Preference Alignment via Principle-Guided Decoding (OPAD) to align model outputs with human preferences during inference.<n>OPAD achieves competitive or superior performance in both general and personalized alignment tasks.
  arXiv  Detail & Related papers  (2025-02-20T02:23:09Z)
• Disentangling Length Bias In Preference Learning Via Response-Conditioned Modeling [87.17041933863041]
  Reinforcement Learning from Human Feedback (RLHF) has achieved considerable success in aligning large language models (LLMs)<n>We introduce a $textbfR$esponse-$textbfc$onditioned $textbfB$radley-$textbfT$erry (Rc-BT) model that enhances the model's capability in length bias mitigating and length instruction following.<n>We also propose the Rc-RM and Rc-DPO algorithm to leverage the Rc-BT model for reward modeling and direct policy optimization
  arXiv  Detail & Related papers  (2025-02-02T14:50:25Z)
• Lifelong Personalized Low-Rank Adaptation of Large Language Models for Recommendation [50.837277466987345]
  We focus on the field of large language models (LLMs) for recommendation. We propose RecLoRA, which incorporates a Personalized LoRA module that maintains independent LoRAs for different users. We also design a Few2Many Learning Strategy, using a conventional recommendation model as a lens to magnify small training spaces to full spaces.
  arXiv  Detail & Related papers  (2024-08-07T04:20:28Z)
• Active Preference Learning for Large Language Models [12.093302163058436]
  We develop an active learning strategy for DPO to make better use of preference labels. We propose a practical acquisition function for prompt/completion pairs based on the predictive entropy of the language model. We demonstrate how our approach improves both the rate of learning and final performance of fine-tuning on pairwise preference data.
  arXiv  Detail & Related papers  (2024-02-12T23:09:00Z)
• Direct Preference Optimization: Your Language Model is Secretly a Reward Model [119.65409513119963]
  We introduce a new parameterization of the reward model in RLHF that enables extraction of the corresponding optimal policy in closed form. The resulting algorithm, which we call Direct Preference Optimization (DPO), is stable, performant, and computationally lightweight. Our experiments show that DPO can fine-tune LMs to align with human preferences as well as or better than existing methods.
  arXiv  Detail & Related papers  (2023-05-29T17:57:46Z)
• Personalizing Intervened Network for Long-tailed Sequential User Behavior Modeling [66.02953670238647]
  Tail users suffer from significantly lower-quality recommendation than the head users after joint training. A model trained on tail users separately still achieve inferior results due to limited data. We propose a novel approach that significantly improves the recommendation performance of the tail users.
  arXiv  Detail & Related papers  (2022-08-19T02:50:19Z)
• Deep Optimized Priors for 3D Shape Modeling and Reconstruction [38.79018852887249]
  We introduce a new learning framework for 3D modeling and reconstruction. We show that the proposed strategy effectively breaks the barriers constrained by the pre-trained priors.
  arXiv  Detail & Related papers  (2020-12-14T03:56:31Z)
• Attention-Based Deep Learning Framework for Human Activity Recognition with User Adaptation [5.629161809575013]
  Sensor-based human activity recognition (HAR) requires to predict the action of a person based on sensor-generated time series data. We propose a novel deep learning framework, algname, based on a purely attention-based mechanism. We show that our proposed attention-based architecture is considerably more powerful than previous approaches.
  arXiv  Detail & Related papers  (2020-06-06T09:26:07Z)

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.