We present a data-driven approach for physics-based, muscle-driven dexterous control that enables musculoskeletal hands to perform precise piano playing for novel pieces of music outside the reference dataset. Our approach combines high-frequency muscle-level control with low-frequency latent-space coordination in a hierarchical architecture. At the low level, general single-hand policies are trained via reinforcement learning to generate dynamic muscle-tendon activations while tracking trajectories from a large reference motion dataset. The resulting tracking policies are then distilled into variational autoencoder (VAE) models, yielding smooth and structured latent spaces that abstract away low-level muscle dynamics. For the high level, we train piece-specific policies to operate in this latent space, coordinating bimanual motions based on specific goals, denoted by note events extracted from given musical scores, to synthesize performances beyond the reference data. High-level control is formulated as a decentralized multiagent reinforcement learning problem combined with adversarial learning for motion imitation. In addition, we present an enhanced musculoskeletal hand model that supports fine control of fingers for accurate low-level motion tracking and diverse high-level motion synthesis. We evaluate the control pipeline of our approach on a diverse piano repertoire spanning multiple musical styles and technical demands. Results demonstrate that our approach can synthesize coordinated bimanual motions with accurate key presses, and achieve the state-of-the-art performance of piano playing in physics-based dexterous control, while generalizing to sheet music that is not presented in the reference dataset. We also show that our musculoskeletal hand model demonstrates superior biomechanical stability and tracking precision compared to the existing model, and validate that our musculoskeletal hand model and muscle-driven controller can generate physiologically plausible activation patterns that align with human electromyography (EMG) recordings when subjects perform multiple tasks.
论文检索
输入标题、作者或关键词，从 3,199 篇学术成果中精准定位
Retargeting motion across characters with varying body shapes while preserving interaction semantics, such as self-contact and near-body proximity, remains a challenging problem. While recent geometry-aware approaches address this by maintaining spatial relationships between predefined corresponding regions, their reliance on static correspondences often struggles when the target character exhibits exaggerated body proportions. In this paper, we present a geometry-aware motion retargeting framework that preserves interaction semantics by performing proximity matching over spatially adaptive anchors. Unlike prior methods with static anchor definitions, the proposed method dynamically repositions anchors to reachable regions on the target character. This is achieved via a Transformer-based anchor refinement strategy that predicts anchor displacements and constrains the translated anchors to remain on the target character geometry through differentiable soft projection. By incorporating pose-dependent spatial structures from the source character, the adapted anchors provide structurally coherent guidance for interaction-aware retargeting. Conditioned on these anchors, a graph-based autoencoder predicts target skeletal motion that preserves the spatial configuration of the source. To encourage task-aligned optimization between anchor adaptation and motion retargeting, we adopt an alternating training scheme in which each module is optimized in turn. Through extensive evaluations, we demonstrate that our method outperforms state-of-the-art approaches in preserving interaction fidelity across diverse character geometries. Code is available at Project Page.
Synthesizing realistic 3D indoor scenes remains challenging due to data scarcity and the difficulty of simultaneously enforcing global architectural constraints and local semantic consistency. Existing approaches often overlook structural boundaries or rely on fully connected relation graphs that introduce redundant generation errors. Inspired by human design cognition, we present CasLayout, a cascaded diffusion framework that decomposes the joint scene generation task into four conditional sub-stages with explicit physical and semantic roles: (1) predicting furniture quantity and categories, (2) refining object sizes and feature embeddings, (3) modeling spatial relationships in a latent space, and (4) generating Oriented Bounding Boxes (OBBs). This decoupled architecture reduces data requirements and enables flexible integration of Large Language Models (LLMs) and Vision Language Models (VLMs) for zero-shot tasks such as image-to-scene generation. To maintain physical validity within complex floor plans, we explicitly model building elements (e.g., walls, doors, and windows) as conditional constraints. Furthermore, to address the high entropy of dense relation graphs, we introduce a sparse relation graph formulation aligned with human spatial descriptions. By encoding these sparse graphs into a compact latent space using a bidirectional Variational Autoencoder (VAE), the proposed framework provides enhanced relational controllability, allowing generated layouts to better respect functional organization. Experiments demonstrate that CasLayout achieves state-of-the-art performance in fidelity and diversity while enabling improved controllability in practical applications.
The fundamental limitation of traditional strand-based modeling is not simply data scarcity, but the ill-posedness of inferring complex 3D fields from 2D imagery without structural constraints. This unconstrained regression leads to catastrophic failures in resolving both global occlusion (e.g., in ponytails) and local directionality (e.g., in curls), resulting in over-smoothed, plausible-but-incorrect geometries. To resolve this, we integrate the strong geometric priors of Large Reconstruction Models (LRMs) into the strand generation pipeline. Using the LRM mesh as a structural anchor, we employ a novel Dual Orientation AutoEncoder to lift coarse geometry into high-fidelity strands. By resolving vector field singularities through latent-space optimization and surface-guided refinement, our method effectively disentangles complex topological structures, setting a new benchmark for robustness and accuracy in hair reconstruction.
Motion warping is a core technique in character animation that enables the adaptation of existing motion data to novel spatio-temporal constraints. Conventional motion warping methods often rely on heuristic modifications that can violate physical consistency or introduce visual artifacts. More recent learning-based editing approaches improve realism, but many of them encode motion into tightly entangled latent space, which makes them struggle to balance editing flexibility and content preservation. To address this, we propose a novel deep motion warping framework that explicitly disentangles the motion structure from global and stylistic attributes for intuitive motion editing. Our key insight is to leverage learned phase features as a continuous and robust representation of the underlying structure, and explicitly disentangle motion into root velocity, phase, and learned latent variables using a phase-conditioned diffusion autoencoder. This design supports a wide range of editing operations, including root motion warping, motion exaggeration, time warping, and style transfer by directly manipulating decoupled components, without requiring paired training data. Extensive experiments demonstrate that our approach enables high-level, flexible motion editing while strictly preserving the structural consistency and physical plausibility of the source motion
Popularity bias is a pervasive challenge in recommender systems, where a few popular items dominate attention while the majority of less popular items remain underexposed. This imbalance can reduce recommendation quality and lead to unfair item exposure. Although existing mitigation methods address this issue to some extent, they often lack transparency in how they operate. In this paper, we propose a post-hoc approach, PopSteer, that leverages a Sparse Autoencoder (SAE) to both interpret and mitigate popularity bias in recommendation models. The SAE is trained to replicate a trained model's behavior while enabling neuron-level interpretability. By introducing synthetic users with strong preferences for either popular or unpopular items, we identify neurons encoding popularity signals through their activation patterns. We then steer recommendations by adjusting the activations of the most biased neurons. Experiments on three public datasets with a sequential recommendation model demonstrate that PopSteer significantly enhances fairness with minimal impact on accuracy, while providing interpretable insights and fine-grained control over the fairness–accuracy trade-off.
Topic modeling plays a critical role in organizing and understanding large-scale web content. While neural topic models (NTMs) based on variational autoencoders (VAEs) have achieved notable success in analyzing textual data, they remain limited in addressing the multimodal nature of modern web content. Existing unimodal or multimodal extensions often suffer from posterior collapse and fail to capture the directional semantics inherent in both text and images, resulting in incoherent topics and limited interpretability. To address these challenges, we propose MM-vNTM (MultiModal Neural Topic Model with von Mises-Fisher Mixtures), a framework for web-scale topic discovery over multimodal data. MM-vNTM leverages pre-aligned cross-modal embeddings as inputs and jointly models document-level representations of text and image modalities in a shared hyperspherical latent space. Furthermore, it defines topics as mixtures of von Mises-Fisher (vMF) distributions in the L2-normalized word embedding space, explicitly capturing directional similarity. Experiments on multimedia web datasets demonstrate that MM-vNTM consistently outperforms state-of-the-art unimodal and multimodal baselines in terms of overall topic quality, highlighting its effectiveness for real-world web scenarios.
Behavioral patterns captured in embeddings learned from interaction data are pivotal across various stages of production recommender systems. However, in the initial retrieval stage, practitioners face an inherent tradeoff between embedding expressiveness and the scalability and latency of serving components, resulting in the need for representations that are both compact and expressive. To address this challenge, we propose a training strategy for learning high-dimensional sparse embedding layers in place of conventional dense ones, balancing efficiency, representational expressiveness, and interpretability. To demonstrate our approach, we modified the production-grade collaborative filtering autoencoder ELSA, achieving up to 10× reduction in embedding size with no loss of recommendation accuracy, and up to 100× reduction with only a 2.5% loss. Moreover, the active embedding dimensions reveal an interpretable inverted-index structure that segments items in a way directly aligned with the model's latent space, thereby enabling integration of segment-level recommendation functionality (e.g., 2D homepage layouts) within the candidate retrieval model itself. Source codes, additional results, as well as a live demo are available at https://github.com/zombak79/compressed\_elsa.
As pretrained models are increasingly shared on the web, ensuring that models can forget or delete sensitive, copyrighted, or private information upon request has become crucial. Machine unlearning has been proposed to address this challenge. However, current evaluations for unlearning methods rely on output-based metrics, which cannot verify whether information is completely deleted or merely suppressed at the representation level, where suppression is insufficient for true unlearning. To address this gap, we propose a novel restoration-based analysis framework that uses Sparse Autoencoders to identify class-specific expert features in intermediate layers and applies inference-time steering to quantitatively distinguish between suppression and deletion. Applying our framework to 12 major unlearning methods in image classification tasks, we find that most methods achieve high restoration rates of unlearned information, indicating that they only suppress information at the decision-boundary level, while preserving semantic features in intermediate representations. Notably, even retraining from pretrained checkpoints shows high restoration, revealing that robust semantic features inherited from pretraining are not removed by retraining. These results demonstrate that representation-level retention poses significant risks overlooked by output-based metrics, highlighting the need for new unlearning evaluation criteria. We propose new evaluation guidelines that prioritize representation-level verification, especially for privacy-critical applications in the era of pre-trained models.
Generative latent diffusion models (LDMs) have been extensively applied in various fields yet underperform in time-series prediction. Therefore, We propose the Re-Diffusion model, a latent diffusion approach that generates backbone residuals specifically tailored for time-series forecasting. The model comprises a variational autoencoder that compresses the residuals between the actual future values and the predictions from the backbone into latent space. It also includes a conditional diffusion generator to forecast the potential distribution of these residuals. Our findings reveal that this latent-space methodology particularly enhances existing backbone predictors, by effectively reducing prediction bias through an advanced estimation of complex error distributions. While previous diffusion-based models tend to struggle with long-term forecasting, Re-Diffusion integrates the strengths of diffusion methods, leading to improvements in long-term predictions. Our experimental results indicate that the Re-Diffusion model achieves a 10% promotion over state-of-art predictors, marking a significant advancement in the field of time-series forecasting.
Multimodal information extraction (MIE) constitutes a set of essential tasks aimed at extracting structural information from Web texts with integrating images, to facilitate the structural construction of Web-based semantic knowledge. To address the expanding category set including newly emerging entity types or relations on websites, prior research proposed the zero-shot MIE (ZS-MIE) task which aims to extract unseen structural knowledge with textual and visual modalities. However, the ZS-MIE models are limited to recognizing the samples that fall within the unseen category set, and they struggle to deal with real-world scenarios that encompass both seen and unseen categories. The shortcomings of existing methods can be ascribed to two main aspects. On one hand, these methods construct representations of samples and categories within Euclidean space, failing to capture the hierarchical semantic relationships between the two modalities within a sample and their corresponding category prototypes. On the other hand, there is a notable gap in the distribution of semantic similarity between seen and unseen category sets, which impacts the generative capability of the ZS-MIE models. To overcome the above disadvantages, we delve into the generalized zero-shot MIE (GZS-MIE) task and propose the hyperbolic multimodal generative representation learning framework (HMGRL). The variational information bottleneck and autoencoder networks are reconstructed with hyperbolic space for modeling the multi-level hierarchical semantic correlations among samples and prototypes. Furthermore, the proposed model is trained with the unseen samples generated by the decoder, and we introduce the semantic similarity distribution alignment loss to enhance the model's generalization performance. Experimental evaluations on two benchmark datasets underscore the superiority of HMGRL compared to existing baseline methods.
Search and Recommendation (S&R) are core information access channels on modern multi-scenario platforms. Existing joint S&R models face two critical challenges: (1) cross-scenario interest inconsistency, failing to unify explicit search intent (queries) and implicit recommendation intent (behavioral interactions) into coherent user interest representations; (2) severe S&R trade-off, where enhancing one task degrades the other due to static knowledge sharing and unbalanced feature utilization. To address these issues, we propose MinSAR , a novel framework focusing on cross-S&R user interest consistency. It integrates two key innovations: a Unified Interest Generation (UIG) module using Vector Quantized-Variational Autoencoder (VQ-VAE) to fuse long-term user preferences (via a user-specific memory network) and dynamic short-term contextual behaviors, generating compact cross-scenario latent representations that bridge explicit and implicit intents. Additionally, an Interest-Guided Attention Expert Network replaces static multi-task gating with intent-aware weight allocation. Guided by UIG's unified interest, it dynamically balances cross-S&R shared knowledge and task-specific expertise (semantic matching for search, collaborative filtering for recommendation), mitigating inter-task conflicts. Extensive experiments on two real-world datasets (KuaiSAR and Amazon Kindle Store) against 13 baselines show MinSAR outperforms state-of-the-art joint S&R models. Further analysis confirms its ability to eliminate the S&R performance trade-off.
Source localization has attracted increasing attention in recent years due to its vital role in governing the harmful propagation. However, existing localization methods do not fully consider the temporal characteristics in propagation and struggle to leverage the continuous-time information of real-world propagation scenarios. Moreover, the aggregation ability of GNN based localization models is limited by the structural noise commonly present in complicated real-world topologies. To address these challenges, a Spectral Neural Ordinary Differential Equation (SNODE) is proposed to infer the source in real-world continuous-time scenarios. First, the forward propagation is formulated as a flow based ODE system, and the source localization problem is transformed into an inverse ODE modeling task. Second, a neural process based on a graph variational autoencoder is introduced to encode global latent propagation patterns as a conditioning variable for the ODE system. Third, a spectral graph optimization is performed to suppress topological noise by filtering out high-frequency components that degrade the quality of graph aggregation in the neural process. Comprehensive experiments demonstrate that SNODE not only outperforms the optimal baseline in real-world continuous-time propagation scenarios with an average performance improvement of 43.8%, but also achieves consistently superior performance in synthetic discrete-time datasets with an improvement of 4.5%, highlighting its strong generalization ability in different propagation settings. Our code is available at https://github.com/cgao-comp/SNODE.
CausalSKyHop: Knowledge-Aware Causal Explanation of Dynamic GNNs via Higher-Order Semantic Reasoning
Dynamic Graph Neural Networks (DyGNNs) are widely used to model web-scale semantic-rich graph data (e.g., social networks, knowledge graphs), but their inability to explain predictions ground-ed in structured knowledge remains a challenge, especially when predictions rely on complex higher-order substructures. We propose CausalSKyHop, a semantic- and knowledge-aware framework that explains DyGNNs by uncovering causal higher-order patterns in evolving knowledge structures. To model the semantic fabric of the graph, CausalSKyHop incorporates a Higher-Order Structural Causal Model to capture multi-node knowledge dependencies, and uses contrastive learning to isolate semantically-meaningful causal relationships from spurious ones. A dynamic correlation module further separates persistent knowledge from evolving semantic contexts. Through knowledge-infused, structure-aware variational graph autoencoders, our method produces interpretable causal subgraphs that capture the dynamic flow of knowledge and semantics. Experimental evaluations on multiple web and knowledge-rich graph benchmarks demonstrate that CausalSKyHop consistently outperforms state-of-the-art explainable DyGNNs, achieving notable improvements in both explanation fidelity and downstream prediction accuracy. A detailed case study further illustrates how our method uncovers stable, semantically coherent causal pathways—in contrast to the fragmented explanations of baseline methods—providing intuitive evidence for its superior interpretability. This work establishes the critical role of explicit semantic and knowledge integration through higher-order causal reasoning for building transparent and trustworthy DyGNNs on the web.
Large Language Model (LLM)-based agentic systems have shown strong capabilities across various tasks. However, existing multi-agent frameworks often rely on static or task-level workflows, which either over-process simple queries or underperform on complex ones, while also neglecting the efficiency-performance trade-offs across heterogeneous LLMs. To address these limitations, we propose Difficulty-Aware Agentic Orchestration (DAAO), which can dynamically generate query-specific multi-agent workflows guided by predicted query difficulty. DAAO comprises three interdependent modules: a variational autoencoder (VAE) for difficulty estimation, a modular operator allocator, and a cost- and performance-aware LLM router. A self-adjusting policy updates difficulty estimates based on workflow success, enabling simpler workflows for easy queries and more complex strategies for harder ones. Experiments on six benchmarks demonstrate that DAAO surpasses prior multi-agent systems in both accuracy and inference efficiency, validating its effectiveness for adaptive, difficulty-aware reasoning. Our code is open-sourced at https://github.com/AutoAgents-ai/DAAO
Graph Masked Autoencoders (GMAEs) have emerged as a notable self-supervised learning approach for graph-structured data. Existing GMAE models primarily focus on reconstructing node-level information, categorizing them as single-scale GMAEs. This methodology, while effective in certain contexts, tends to overlook the complex hierarchical structures inherent in many real-world graphs. For instance, molecular graphs exhibit a clear hierarchical organization in the form of the atoms-functional groups-molecules structure. Therefore, the inability of single-scale GMAE models to incorporate these hierarchical relationships often results in an inadequate capture of crucial high-level graph information, leading to a noticeable decline in performance. To address this limitation, we propose Hierarchical Graph Masked AutoEncoders (Hi-GMAE), a novel multi-scale GMAE framework designed to handle the hierarchical structures within graphs. First, Hi-GMAE constructs a multi-scale graph hierarchy through graph pooling, enabling the exploration of graph structures across different granularity levels. To ensure masking uniformity of subgraphs across these scales, we propose a novel coarse-to-fine strategy that initiates masking at the coarsest scale and progressively back-projects the mask to finer scales. Furthermore, we integrate a gradual recovery strategy with the masking process to mitigate the learning challenges posed by completely masked subgraphs. Diverging from the standard graph neural network (GNN) used in GMAE models, Hi-GMAE modifies its encoder and decoder into hierarchical structures. This entails using GNN at the finer scales for detailed local graph analysis and employing a graph transformer at coarser scales to capture global information. Such a design enables Hi-GMAE to effectively capture the multi-level information inherent in complex graph structures. Our experiments on 17 graph datasets, covering two graph learning tasks, consistently demonstrate that Hi-GMAE outperforms 29 state-of-the-art self-supervised competitors in capturing comprehensive graph information.
Graph self-supervised learning (GSSL) alleviates the graph data labeling bottleneck without supervision, enabling wide application in domains like recommendation systems and social network analysis. High-frequency signals are valuable in GSSL for capturing local structural preferences, thereby enriching graph representations and boosting model performance. However, in practical applications, two critical problems hinder the efficient and robust use of these signals. First, the locality of high-frequency signals limits their full utilization by the model. Second, over-reliance on specific high-frequency signals will affect the model's generalization. To address the above problems, we propose the Frequency-Corrupt Based Graph Self-Supervised Learning (FC-GSSL) algorithm. Specifically, we generate corrupted graphs biased toward high-frequency signals by corrupting nodes and edges according to their low-frequency contributions. These corrupted graphs are fed as input to an autoencoder, with low-frequency and general features serving as the supervision. This compels the model to effectively fuse high- and low-frequency signals, thereby integrating and utilizing more valuable high-frequency information. Additionally, we design multiple sampling strategies and form diverse corrupted graphs based on the intersections and union between the results obtained from these strategies. By aligning the node representations from these views, the model can identify valuable frequency combinations, which helps reduce the negative impact of specific high-frequency components and improve generalization. FC-GSSL optimizes the design of GSSL for web applications, significantly improving model performance on complex web-related graphs, such as social networks and citation networks. This work makes a direct contribution to advancing the ''Graph Algorithms and Modeling for the Web'' research track. Experimental results on 14 datasets across multiple tasks illustrate the superiority of the proposed approach.
Subgraph Anomaly Detection (SAD) is crucial for identifying groups that deviate from the regular pattern within graphs, which benefits different domains such as financial fraud and network security. However, current studies rely on traditional node detection methods and fixed sampling strategies of subgraph structures, which makes it difficult to learn the pattern collaboration behavior of subgraphs. To address this limitation, this paper proposes a novel unsupervised framework named PC-SAD. The PC-SAD framework first employs an improved Graph AutoEncoder to identify core anomaly nodes by capturing multi-scale neighborhood information. Starting from these core anomaly nodes, we sample candidate subgraphs with path, tree, and cyclic structures, and enhance them according to the characteristics of the subgraph structures. Subsequently, candidate subgraphs are fed into the proposed Pattern Collaboration-based Graph Contrastive Learning method to generate collaborative pattern embeddings, thereby distinguishing anomaly subgraphs. The experimental results show that PC-SAD outperforms the state-of-the-art baseline methods on four benchmark datasets, which proves that PC-SAD is an effective solution to detect anomaly subgraphs.
Venous thromboembolism (VTE) is a common and life-threatening complication in cancer patients after treatment. Early risk assessment and detection of VTE primarily rely on clinical indicators, such as blood test results. However, existing studies are limited to static or snapshot-based models, failing to capture the evolving dynamics of disease progression, as deep time-series modeling is hindered by the lack of longitudinal clinical data. To address this gap, we introduce CliTsVTE, a large-scale clinical time-series dataset curated for VTE modeling, comprising 501,063 samples from 26,022 patients over seven years across nine cancer types. The dataset contains continuous time gaps between consecutive time points. Unlike many benchmarks, CliTsVTE reflects real-world clinical settings and presents unique challenges in continuous irregular time-series modeling with long-term irregularity and varying data granularity, which makes missingness significantly consequential. To tackle this, we propose a deep learning framework integrating multiple sequential backbones with an adversarially regularized autoencoder (ARAE) that learns latent representations to eliminate missingness. Experiments on CliTsVTE show that our best model achieves 88.7% accuracy and an AUC of 0.952, significantly outperforming traditional time-point models and regular time-series benchmarks. These results establish a strong benchmark for deep modeling of continuous irregularity in clinical time-series data and highlight the potential of AI-driven large-scale clinical datasets in solving real-world medical research challenges.
The increasing frequency and severity of climate-related disasters have intensified the need for real-time monitoring, early warning, and informed decision-making. Earth Observation (EO), powered by satellite data and Machine Learning (ML), offers powerful tools to meet these challenges. Foundation Models (FMs) have revolutionized EO ML by enabling general-purpose pretraining on large-scale remote sensing datasets. However most existing models rely on high-resolution satellite imagery with low revisit rates, limiting their suitability for fast-evolving phenomena and time-critical emergency response. In this work, we present HighFM, a first cut approach towards a FM for high-temporal-resolution, multispectral EO data. Leveraging over 2 TB of SEVIRI imagery from the Meteosat Second Generation (MSG) platform, we adapt the SatMAE masked autoencoding framework to learn robust spatiotemporal representations. To support real-time monitoring, we enhance the original architecture with fine-grained temporal encodings to capture short-term variability. The pretrained models are then fine-tuned on cloud masking and active fire detection tasks. We benchmark our SEVIRI-pretrained Vision Transformers against traditional baselines and recent geospatial FMs, demonstrating consistent gains across both balanced accuracy and IoU metrics. Our results highlight the potential of temporally dense geostationary data for real-time EO, offering a scalable path toward foundation models for disaster detection and tracking.