Search Results for author:
Found 114 papers, 43 papers with code
A Survey of Interactive Generative Video
Interactive Generative Video (IGV) has emerged as a crucial technology in response to the growing demand for high-quality, interactive video content across various domains.
FIM: Frequency-Aware Multi-View Interest Modeling for Local-Life Service Recommendation
There are two main challenges in modeling users' periodic behaviors in the local-life service recommendation systems: 1) the diverse demands of users exhibit varying periodicities, which are difficult to distinguish as they are mixed in the behavior sequences; 2) the periodic behaviors of users are subject to dynamic changes due to factors such as holidays and promotional events.
Comprehensive List Generation for Multi-Generator Reranking
Reranking models solve the final recommendation lists that best fulfill users' demands.
VLM as Policy: Common-Law Content Moderation Framework for Short Video Platform
Offline experiments and large-scale online A/B test demonstrates the superiority of KuaiMod: KuaiMod achieves the best moderation performance on our benchmark.
Generative Auto-Bidding with Value-Guided Explorations
Auto-bidding, with its strong capability to optimize bidding decisions within dynamic and competitive online environments, has become a pivotal strategy for advertising platforms.
BBQRec: Behavior-Bind Quantization for Multi-Modal Sequential Recommendation
Multi-modal sequential recommendation systems leverage auxiliary signals (e. g., text, images) to alleviate data sparsity in user-item interactions.
CHIME: A Compressive Framework for Holistic Interest Modeling
CHIME demonstrates superior ranking performance across diverse datasets, establishing a robust solution for scalable holistic interest modeling in recommendation systems.
LLM-Alignment Live-Streaming Recommendation
In recent years, integrated short-video and live-streaming platforms have gained massive global adoption, offering dynamic content creation and consumption.
Any2Caption:Interpreting Any Condition to Caption for Controllable Video Generation
To address the bottleneck of accurate user intent interpretation within the current video generation community, we present Any2Caption, a novel framework for controllable video generation under any condition.
FullDiT: Multi-Task Video Generative Foundation Model with Full Attention
Current video generative foundation models primarily focus on text-to-video tasks, providing limited control for fine-grained video content creation.
Boosting Resolution Generalization of Diffusion Transformers with Randomized Positional Encodings
In this paper, we propose a novel two-dimensional randomized positional encodings (RPE-2D) framework that focuses on learning positional order of image patches instead of the specific distances between them, enabling seamless high- and low-resolution image generation without requiring high- and low-resolution image training.
DiffMoE: Dynamic Token Selection for Scalable Diffusion Transformers
Diffusion models have demonstrated remarkable success in various image generation tasks, but their performance is often limited by the uniform processing of inputs across varying conditions and noise levels.
Ranked #15 on
Text-to-Image Generation
on GenEval
Learning Cascade Ranking as One Network
Recent advances such as RankFlow and FS-LTR have introduced interaction-aware training paradigms but still struggle to 1) align training objectives with the goal of the entire cascade ranking (i. e., end-to-end recall) and 2) learn effective collaboration patterns for different stages.
Creator-Side Recommender System: Challenges, Designs, and Applications
To this end, we develop a creator-side recommender system, called DualRec, to answer the following question: how to find the most suitable users for each item to enhance the creators' experience?
DSMoE: Matrix-Partitioned Experts with Dynamic Routing for Computation-Efficient Dense LLMs
As large language models continue to scale, computational costs and resource consumption have emerged as significant challenges.
From Principles to Applications: A Comprehensive Survey of Discrete Tokenizers in Generation, Comprehension, Recommendation, and Information Retrieval
Discrete tokenizers have emerged as indispensable components in modern machine learning systems, particularly within the context of autoregressive modeling and large language models (LLMs).
Finedeep: Mitigating Sparse Activation in Dense LLMs via Multi-Layer Fine-Grained Experts
Moreover, we find that Finedeep achieves optimal results when balancing depth and width, specifically by adjusting the number of expert sub-layers and the number of experts per sub-layer.
CineMaster: A 3D-Aware and Controllable Framework for Cinematic Text-to-Video Generation
In the first stage, we design an interactive workflow that allows users to intuitively construct 3D-aware conditional signals by positioning object bounding boxes and defining camera movements within the 3D space.
ChorusCVR: Chorus Supervision for Entire Space Post-Click Conversion Rate Modeling
Post-click conversion rate (CVR) estimation is a vital task in many recommender systems of revenue businesses, e. g., e-commerce and advertising.
Improving Video Generation with Human Feedback
Video generation has achieved significant advances through rectified flow techniques, but issues like unsmooth motion and misalignment between videos and prompts persist.
ConceptMaster: Multi-Concept Video Customization on Diffusion Transformer Models Without Test-Time Tuning
We identify two key challenges for this task: 1) the identity decoupling issue, where directly adopting existing customization methods inevitably mix identity attributes when handling multiple concepts simultaneously, and 2) the scarcity of high-quality video-entity pairs, which is crucial for training a model that can well represent and decouple various customized concepts in video generation.
CRM: Retrieval Model with Controllable Condition
This modification enables the retrieval stage could fulfill the target gap with ranking model, enhancing the retrieval model ability to search item candidates satisfied the user interests and condition effectively.
Adaptive$^2$: Adaptive Domain Mining for Fine-grained Domain Adaptation Modeling
To the best of our knowledge, Adaptive$^2$ is the first approach to automatically learn both domain identification and adaptation in online advertising, opening new research directions for this area.
Towards Precise Scaling Laws for Video Diffusion Transformers
Achieving optimal performance of video diffusion transformers within given data and compute budget is crucial due to their high training costs.
Enhancing Instruction-Following Capability of Visual-Language Models by Reducing Image Redundancy
In this paper, we propose Visual-Modality Token Compression (VMTC) and Cross-Modality Attention Inhibition (CMAI) strategies to alleviate this gap between MLLMs and LLMs by inhibiting the influence of irrelevant visual tokens during content generation, increasing the instruction-following ability of the MLLMs while retaining their multimodal understanding capacity.
Scaling Laws for Online Advertisement Retrieval
To the best of our knowledge, this is the first work to study the scaling laws for online advertisement retrieval of real-world systems, showing great potential for scaling law in advertising system optimization.
QARM: Quantitative Alignment Multi-Modal Recommendation at Kuaishou
In recent years, with the significant evolution of multi-modal large models, many recommender researchers realized the potential of multi-modal information for user interest modeling.
KuaiFormer: Transformer-Based Retrieval at Kuaishou
In large-scale content recommendation systems, retrieval serves as the initial stage in the pipeline, responsible for selecting thousands of candidate items from billions of options to pass on to ranking modules.
MARM: Unlocking the Future of Recommendation Systems through Memory Augmentation and Scalable Complexity
Considering the above differences with LLM, we can draw a conclusion that: for a RecSys model, compared to model parameters, the computational complexity FLOPs is a more expensive factor that requires careful control.
RecFlow: An Industrial Full Flow Recommendation Dataset
Industrial recommendation systems (RS) rely on the multi-stage pipeline to balance effectiveness and efficiency when delivering items from a vast corpus to users.
Coarse-to-fine Dynamic Uplift Modeling for Real-time Video Recommendation
These two modules work together to dynamically identify and targeting specific user groups and applying treatments effectively.
ERABAL: Enhancing Role-Playing Agents through Boundary-Aware Learning
Role-playing is an emerging application in the field of Human-Computer Interaction (HCI), primarily implemented through the alignment training of a large language model (LLM) with assigned characters.
RPAF: A Reinforcement Prediction-Allocation Framework for Cache Allocation in Large-Scale Recommender Systems
Then, we propose a reinforcement prediction-allocation framework (RPAF) to address these issues.
HoME: Hierarchy of Multi-Gate Experts for Multi-Task Learning at Kuaishou
(3) Expert Underfitting: In our services, we have dozens of behavior tasks that need to be predicted, but we find that some data-sparse prediction tasks tend to ignore their specific-experts and assign large weights to shared-experts.
TWIN V2: Scaling Ultra-Long User Behavior Sequence Modeling for Enhanced CTR Prediction at Kuaishou
The significance of modeling long-term user interests for CTR prediction tasks in large-scale recommendation systems is progressively gaining attention among researchers and practitioners.
Multi-Epoch learning with Data Augmentation for Deep Click-Through Rate Prediction
MEDA minimizes overfitting by reducing the dependency of the embedding layer on subsequent training data or the Multi-Layer Perceptron (MLP) layers, and achieves data augmentation through training the MLP with varied embedding spaces.
Small Agent Can Also Rock! Empowering Small Language Models as Hallucination Detector
Hallucination detection is a challenging task for large language models (LLMs), and existing studies heavily rely on powerful closed-source LLMs such as GPT-4.
Decoding at the Speed of Thought: Harnessing Parallel Decoding of Lexical Units for LLMs
The core of our approach is the observation that a pre-trained language model can confidently predict multiple contiguous tokens, forming the basis for a \textit{lexical unit}, in which these contiguous tokens could be decoded in parallel.
RectifID: Personalizing Rectified Flow with Anchored Classifier Guidance
Our study shows that based on a recent rectified flow framework, the major limitation of vanilla classifier guidance in requiring a special classifier can be resolved with a simple fixed-point solution, allowing flexible personalization with off-the-shelf image discriminators.
Modeling User Fatigue for Sequential Recommendation
In this paper, we propose to model user Fatigue in interest learning for sequential Recommendations (FRec).
Full Stage Learning to Rank: A Unified Framework for Multi-Stage Systems
To address this issue, we propose an improved ranking principle for multi-stage systems, namely the Generalized Probability Ranking Principle (GPRP), to emphasize both the selection bias in each stage of the system pipeline as well as the underlying interest of users.
LEARN: Knowledge Adaptation from Large Language Model to Recommendation for Practical Industrial Application
Contemporary recommendation systems predominantly rely on ID embedding to capture latent associations among users and items.
M3oE: Multi-Domain Multi-Task Mixture-of Experts Recommendation Framework
M3oE integrates multi-domain information, maps knowledge across domains and tasks, and optimizes multiple objectives.
Inductive-Deductive Strategy Reuse for Multi-Turn Instructional Dialogues
In this paper, we propose to explicitly capture the complex rules to help the user simulator pose diverse and in-depth instruction.
RecGPT: Generative Personalized Prompts for Sequential Recommendation via ChatGPT Training Paradigm
For the model part, we adopt Generative Pre-training Transformer (GPT) as the sequential recommendation model and design a user modular to capture personalized information.
Sequential Recommendation for Optimizing Both Immediate Feedback and Long-term Retention
In the landscape of Recommender System (RS) applications, reinforcement learning (RL) has recently emerged as a powerful tool, primarily due to its proficiency in optimizing long-term rewards.
Enhancing Role-playing Systems through Aggressive Queries: Evaluation and Improvement
Experiments on models improved by RoleAD indicate that our adversarial dataset ameliorates this deficiency, with the improvements demonstrating a degree of generalizability in ordinary scenarios.
Video-LaVIT: Unified Video-Language Pre-training with Decoupled Visual-Motional Tokenization
In light of recent advances in multimodal Large Language Models (LLMs), there is increasing attention to scaling them from image-text data to more informative real-world videos.
Ranked #3 on
Text-to-Video Generation
on MSR-VTT
Future Impact Decomposition in Request-level Recommendations
Furthermore, we show that a reward-based future decomposition strategy can better express the item-wise future impact and improve the recommendation accuracy in the long term.
Just Ask One More Time! Self-Agreement Improves Reasoning of Language Models in (Almost) All Scenarios
Although chain-of-thought (CoT) prompting combined with language models has achieved encouraging results on complex reasoning tasks, the naive greedy decoding used in CoT prompting usually causes the repetitiveness and local optimality.
Mixed Attention Network for Cross-domain Sequential Recommendation
Recent proposed cross-domain sequential recommendation models such as PiNet and DASL have a common drawback relying heavily on overlapped users in different domains, which limits their usage in practical recommender systems.
Inverse Learning with Extremely Sparse Feedback for Recommendation
In this paper, we propose a meta-learning method to annotate the unlabeled data from loss and gradient perspectives, which considers the noises in both positive and negative instances.
DialogBench: Evaluating LLMs as Human-like Dialogue Systems
In this paper, we propose DialogBench, a dialogue evaluation benchmark that contains 12 dialogue tasks to probe the capabilities of LLMs as human-like dialogue systems should have.
Adaptive Neural Ranking Framework: Toward Maximized Business Goal for Cascade Ranking Systems
Concretely, we employ multi-task learning to adaptively combine the optimization of relaxed and full targets, which refers to metrics Recall@m@k and OPA respectively.
KwaiYiiMath: Technical Report
Recent advancements in large language models (LLMs) have demonstrated remarkable abilities in handling a variety of natural language processing (NLP) downstream tasks, even on mathematical tasks requiring multi-step reasoning.
Ranked #97 on
Arithmetic Reasoning
on GSM8K
(using extra training data)
Parrot: Enhancing Multi-Turn Instruction Following for Large Language Models
In this paper, we introduce Parrot, a solution aiming to enhance multi-turn instruction following for LLMs.
AURO: Reinforcement Learning for Adaptive User Retention Optimization in Recommender Systems
As the policy performance of RL is sensitive to environment drifts, the loss function enables the state abstraction to be reflective of environment changes and notify the recommendation policy to adapt accordingly.
Generative Retrieval with Semantic Tree-Structured Item Identifiers via Contrastive Learning
To obtain efficiency and effectiveness, this paper introduces a generative retrieval framework, namely SEATER, which learns SEmAntic Tree-structured item identifiERs via contrastive learning.
KuaiSim: A Comprehensive Simulator for Recommender Systems
For each task, KuaiSim also provides evaluation protocols and baseline recommendation algorithms that further serve as benchmarks for future research.
Unified Language-Vision Pretraining in LLM with Dynamic Discrete Visual Tokenization
Specifically, we introduce a well-designed visual tokenizer to translate the non-linguistic image into a sequence of discrete tokens like a foreign language that LLM can read.
A Large Language Model Enhanced Conversational Recommender System
For response generation, we utilize the generation ability of LLM as a language interface to better interact with users.
Understanding and Modeling Passive-Negative Feedback for Short-video Sequential Recommendation
To enhance the robustness of our model, we then introduce a multi-task learning module to simultaneously optimize two kinds of feedback -- passive-negative feedback and traditional randomly-sampled negative feedback.
Graph Contrastive Learning with Generative Adversarial Network
Specifically, we present GACN, a novel Generative Adversarial Contrastive learning Network for graph representation learning.
PANE-GNN: Unifying Positive and Negative Edges in Graph Neural Networks for Recommendation
Recommender systems play a crucial role in addressing the issue of information overload by delivering personalized recommendations to users.
State Regularized Policy Optimization on Data with Dynamics Shift
Data with dynamics shift are separated according to their environment parameters to train the corresponding policy.
Generative Flow Network for Listwise Recommendation
In this work, we aim to learn a policy that can generate sufficiently diverse item lists for users while maintaining high recommendation quality.
Multi-behavior Self-supervised Learning for Recommendation
Secondly, existing methods utilizing self-supervised learning (SSL) to tackle the data sparsity neglect the serious optimization imbalance between the SSL task and the target task.
When Search Meets Recommendation: Learning Disentangled Search Representation for Recommendation
In our paper, we propose a Search-Enhanced framework for the Sequential Recommendation (SESRec) that leverages users' search interests for recommendation, by disentangling similar and dissimilar representations within S&R behaviors.
Exploration and Regularization of the Latent Action Space in Recommendation
To overcome this challenge, we propose a hyper-actor and critic learning framework where the policy decomposes the item list generation process into a hyper-action inference step and an effect-action selection step.
Multi-Task Recommendations with Reinforcement Learning
To be specific, the RMTL structure can address the two aforementioned issues by (i) constructing an MTL environment from session-wise interactions and (ii) training multi-task actor-critic network structure, which is compatible with most existing MTL-based recommendation models, and (iii) optimizing and fine-tuning the MTL loss function using the weights generated by critic networks.
Disentangled Causal Embedding With Contrastive Learning For Recommender System
Most previous studies neglect user's conformity and entangle interest with it, which may cause the recommender systems fail to provide satisfying results.
Divide and Conquer: Towards Better Embedding-based Retrieval for Recommender Systems From a Multi-task Perspective
Embedding-based retrieval (EBR) methods are widely used in modern recommender systems thanks to its simplicity and effectiveness.
TWIN: TWo-stage Interest Network for Lifelong User Behavior Modeling in CTR Prediction at Kuaishou
And for the user-item cross features, we compress each into a one-dimentional bias term in the attention score calculation to save the computational cost.
Two-Stage Constrained Actor-Critic for Short Video Recommendation
One the one hand, the platforms aims at optimizing the users' cumulative watch time (main goal) in long term, which can be effectively optimized by Reinforcement Learning.
Reinforcing User Retention in a Billion Scale Short Video Recommender System
In this paper, we choose reinforcement learning methods to optimize the retention as they are designed to maximize the long-term performance.
PEPNet: Parameter and Embedding Personalized Network for Infusing with Personalized Prior Information
By infusing personalized selection of Embedding and personalized modification of DNN parameters, PEPNet tailored to the interests of each individual obtains significant performance gains, with online improvements exceeding 1\% in multiple task metrics across multiple domains.
PrefRec: Recommender Systems with Human Preferences for Reinforcing Long-term User Engagement
Though promising, the application of RL heavily relies on well-designed rewards, but designing rewards related to long-term user engagement is quite difficult.
Real-time Short Video Recommendation on Mobile Devices
However, as users continue to watch videos and feedback, the changing context leads the ranking of the server-side recommendation system inaccurate.
ResAct: Reinforcing Long-term Engagement in Sequential Recommendation with Residual Actor
Meanwhile, reinforcement learning (RL) is widely regarded as a promising framework for optimizing long-term engagement in sequential recommendation.
Truncation-Free Matching System for Display Advertising at Alibaba
When receiving a user request, matching system (i) finds the crowds that the user belongs to; (ii) retrieves all ads that have targeted those crowds.
Optimizing Multiple Performance Metrics with Deep GSP Auctions for E-commerce Advertising
In e-commerce advertising, the ad platform usually relies on auction mechanisms to optimize different performance metrics, such as user experience, advertiser utility, and platform revenue.
Exploration in Online Advertising Systems with Deep Uncertainty-Aware Learning
In this paper, we propose a novel Deep Uncertainty-Aware Learning (DUAL) method to learn CTR models based on Gaussian processes, which can provide predictive uncertainty estimations while maintaining the flexibility of deep neural networks.
Learning to Infer User Hidden States for Online Sequential Advertising
To drive purchase in online advertising, it is of the advertiser's great interest to optimize the sequential advertising strategy whose performance and interpretability are both important.
A Deep Prediction Network for Understanding Advertiser Intent and Satisfaction
For e-commerce platforms such as Taobao and Amazon, advertisers play an important role in the entire digital ecosystem: their behaviors explicitly influence users' browsing and shopping experience; more importantly, advertiser's expenditure on advertising constitutes a primary source of platform revenue.
COLD: Towards the Next Generation of Pre-Ranking System
We name it COLD (Computing power cost-aware Online and Lightweight Deep pre-ranking system).
Dynamic Knapsack Optimization Towards Efficient Multi-Channel Sequential Advertising
In E-commerce, advertising is essential for merchants to reach their target users.
Learning Optimal Tree Models Under Beam Search
Retrieving relevant targets from an extremely large target set under computational limits is a common challenge for information retrieval and recommendation systems.
DCAF: A Dynamic Computation Allocation Framework for Online Serving System
These stages usually allocate resource manually with specific computing power budgets, which requires the serving configuration to adapt accordingly.
Search-based User Interest Modeling with Lifelong Sequential Behavior Data for Click-Through Rate Prediction
Serving the main traffic in our real system now, SIM models user behavior data with maximum length reaching up to 54000, pushing SOTA to 54x.
Learning to Accelerate Heuristic Searching for Large-Scale Maximum Weighted b-Matching Problems in Online Advertising
Bipartite b-matching is fundamental in algorithm design, and has been widely applied into economic markets, labor markets, etc.
A Deep Recurrent Survival Model for Unbiased Ranking
Position bias is a critical problem in information retrieval when dealing with implicit yet biased user feedback data.
Progressive Feature Polishing Network for Salient Object Detection
Feature matters for salient object detection.
Stick to the Facts: Learning towards a Fidelity-oriented E-Commerce Product Description Generation
Different from other text generation tasks, in product description generation, it is of vital importance to generate faithful descriptions that stick to the product attribute information.
Learning to Advertise for Organic Traffic Maximization in E-Commerce Product Feeds
We refer to this process as Leverage.
Res-embedding for Deep Learning Based Click-Through Rate Prediction Modeling
In this paper, we model user behavior using an interest delay model, study carefully the embedding mechanism, and obtain two important results: (i) We theoretically prove that small aggregation radius of embedding vectors of items which belongs to a same user interest domain will result in good generalization performance of deep CTR model.
Practice on Long Sequential User Behavior Modeling for Click-Through Rate Prediction
To our knowledge, this is one of the first industrial solutions that are capable of handling long sequential user behavior data with length scaling up to thousands.
Lifelong Sequential Modeling with Personalized Memorization for User Response Prediction
In order to tackle these challenges, in this paper, we propose a Hierarchical Periodic Memory Network for lifelong sequential modeling with personalized memorization of sequential patterns for each user.
Joint Optimization of Tree-based Index and Deep Model for Recommender Systems
The previous work Tree-based Deep Model (TDM) \cite{zhu2018learning} greatly improves recommendation accuracy using tree index.
Deep Interest Evolution Network for Click-Through Rate Prediction
Easy-to-use, Modular and Extendible package of deep-learning based CTR models. DeepFM, DeepInterestNetwork(DIN), DeepInterestEvolutionNetwork(DIEN), DeepCrossNetwork(DCN), AttentionalFactorizationMachine(AFM), Neural Factorization Machine(NFM), AutoInt
Ranked #1 on
Click-Through Rate Prediction
on Amazon Dataset
Learning Adaptive Display Exposure for Real-Time Advertising
In this paper, we investigate the problem of advertising with adaptive exposure: can we dynamically determine the number and positions of ads for each user visit under certain business constraints so that the platform revenue can be increased?
A Multi-Agent Reinforcement Learning Method for Impression Allocation in Online Display Advertising
With this formulation, we derive the optimal impression allocation strategy by solving the optimal bidding functions for contracts.
Multi-agent Reinforcement Learning
reinforcement-learning
+2
Semantic Human Matting
SHM is the first algorithm that learns to jointly fit both semantic information and high quality details with deep networks.
Ranked #5 on
Image Matting
on AIM-500
Entire Space Multi-Task Model: An Effective Approach for Estimating Post-Click Conversion Rate
To the best of our knowledge, this is the first public dataset which contains samples with sequential dependence of click and conversion labels for CVR modeling.
Real-Time Bidding with Multi-Agent Reinforcement Learning in Display Advertising
Real-time advertising allows advertisers to bid for each impression for a visiting user.
Budget Constrained Bidding by Model-free Reinforcement Learning in Display Advertising
Our analysis shows that the immediate reward from environment is misleading under a critical resource constraint.
Learning Tree-based Deep Model for Recommender Systems
In systems with large corpus, however, the calculation cost for the learnt model to predict all user-item preferences is tremendous, which makes full corpus retrieval extremely difficult.
Image Matters: Visually modeling user behaviors using Advanced Model Server
So we propose to model user preference jointly with user behavior ID features and behavior images.
Rocket Launching: A Universal and Efficient Framework for Training Well-performing Light Net
Models applied on real time response task, like click-through rate (CTR) prediction model, require high accuracy and rigorous response time.
Deep Interest Network for Click-Through Rate Prediction
In this way, user features are compressed into a fixed-length representation vector, in regardless of what candidate ads are.
Ranked #1 on
Click-Through Rate Prediction
on Amazon
Learning Piece-wise Linear Models from Large Scale Data for Ad Click Prediction
CTR prediction in real-world business is a difficult machine learning problem with large scale nonlinear sparse data.
Optimized Cost per Click in Taobao Display Advertising
Moreover, the platform has to be responsible for the business revenue and user experience.
Learning Kernels with Radiuses of Minimum Enclosing Balls
Experiments show that our method significantly outperforms both SVM with the uniform combination of basis kernels and other state-of-art MKL approaches.
