arXiv subjects

Kun Zhang

Publications and source records attributed to Kun Zhang.

At least 217 records · Page 12Linked to original sources

arXiv2025

Causal Representation Learning from Multimodal Biomedical Observations

Prevalent in biomedical applications (e.g., human phenotype research), multimodal datasets can provide valuable insights into the underlying physiological mechanisms. However, current machine learning (ML) models designed to analyze these datasets often lack interpretability and identifiability guarantees, which are essential for biomedical research. Recent advances in causal representation learning have shown promise in identifying interpretable latent causal variables with formal theoretical guarantees. Unfortunately, most current work on multimodal distributions either relies on restrictive parametric assumptions or yields only coarse identification results, limiting their applicability to biomedical research that favors a detailed understanding of the mechanisms. In this work, we aim to develop flexible identification conditions for multimodal data and principled methods to facilitate the understanding of biomedical datasets. Theoretically, we consider a nonparametric latent distribution (c.f., parametric assumptions in previous work) that allows for causal relationships across potentially different modalities. We establish identifiability guarantees for each latent component, extending the subspace identification results from previous work. Our key theoretical contribution is the structural sparsity of causal connections between modalities, which, as we will discuss, is natural for a large collection of biomedical systems. Empirically, we present a practical framework to instantiate our theoretical insights. We demonstrate the effectiveness of our approach through extensive experiments on both numerical and synthetic datasets. Results on a real-world human phenotype dataset are consistent with established biomedical research, validating our theoretical and methodological framework.

cs.LG↗

arXiv2025

Type Information-Assisted Self-Supervised Knowledge Graph Denoising

Knowledge graphs serve as critical resources supporting intelligent systems, but they can be noisy due to imperfect automatic generation processes. Existing approaches to noise detection often rely on external facts, logical rule constraints, or structural embeddings. These methods are often challenged by imperfect entity alignment, flexible knowledge graph construction, and overfitting on structures. In this paper, we propose to exploit the consistency between entity and relation type information for noise detection, resulting a novel self-supervised knowledge graph denoising method that avoids those problems. We formalize type inconsistency noise as triples that deviate from the majority with respect to type-dependent reasoning along the topological structure. Specifically, we first extract a compact representation of a given knowledge graph via an encoder that models the type dependencies of triples. Then, the decoder reconstructs the original input knowledge graph based on the compact representation. It is worth noting that, our proposal has the potential to address the problems of knowledge graph compression and completion, although this is not our focus. For the specific task of noise detection, the discrepancy between the reconstruction results and the input knowledge graph provides an opportunity for denoising, which is facilitated by the type consistency embedded in our method. Experimental validation demonstrates the effectiveness of our approach in detecting potential noise in real-world data.

cs.LG↗

arXiv2025

SDD-4DGS: Static-Dynamic Aware Decoupling in Gaussian Splatting for 4D Scene Reconstruction

Dynamic and static components in scenes often exhibit distinct properties, yet most 4D reconstruction methods treat them indiscriminately, leading to suboptimal performance in both cases. This work introduces SDD-4DGS, the first framework for static-dynamic decoupled 4D scene reconstruction based on Gaussian Splatting. Our approach is built upon a novel probabilistic dynamic perception coefficient that is naturally integrated into the Gaussian reconstruction pipeline, enabling adaptive separation of static and dynamic components. With carefully designed implementation strategies to realize this theoretical framework, our method effectively facilitates explicit learning of motion patterns for dynamic elements while maintaining geometric stability for static structures. Extensive experiments on five benchmark datasets demonstrate that SDD-4DGS consistently outperforms state-of-the-art methods in reconstruction fidelity, with enhanced detail restoration for static structures and precise modeling of dynamic motions. The code will be released.

cs.CV↗

arXiv2025

Generative Artificial Intelligence in Robotic Manipulation: A Survey

This survey provides a comprehensive review on recent advancements of generative learning models in robotic manipulation, addressing key challenges in the field. Robotic manipulation faces critical bottlenecks, including significant challenges in insufficient data and inefficient data acquisition, long-horizon and complex task planning, and the multi-modality reasoning ability for robust policy learning performance across diverse environments. To tackle these challenges, this survey introduces several generative model paradigms, including Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), diffusion models, probabilistic flow models, and autoregressive models, highlighting their strengths and limitations. The applications of these models are categorized into three hierarchical layers: the Foundation Layer, focusing on data generation and reward generation; the Intermediate Layer, covering language, code, visual, and state generation; and the Policy Layer, emphasizing grasp generation and trajectory generation. Each layer is explored in detail, along with notable works that have advanced the state of the art. Finally, the survey outlines future research directions and challenges, emphasizing the need for improved efficiency in data utilization, better handling of long-horizon tasks, and enhanced generalization across diverse robotic scenarios. All the related resources, including research papers, open-source data, and projects, are collected for the community in https://github.com/GAI4Manipulation/AwesomeGAIManipulation

cs.RO↗

arXiv2025

Nonstationary Time Series Forecasting via Unknown Distribution Adaptation

As environments evolve, temporal distribution shifts can degrade time series forecasting performance. A straightforward solution is to adapt to nonstationary changes while preserving stationary dependencies. Hence, some methods disentangle stationary and nonstationary components by assuming uniform distribution shifts, but it is impractical since when the distribution changes is unknown. To address this challenge, we propose the \textbf{U}nknown \textbf{D}istribution \textbf{A}daptation (\textbf{UDA}) model for nonstationary time series forecasting, which detects when distribution shifts occur and disentangles stationary/nonstationary latent variables, thus enabling adaptation to unknown distribution without assuming a uniform distribution shift. Specifically, under a Hidden Markov assumption of latent environments, we demonstrate that the latent environments are identifiable. Sequentially, we further disentangle stationary/nonstationary latent variables by leveraging the variability of historical information. Based on these theoretical results, we propose a variational autoencoder-based model, which incorporates an autoregressive hidden Markov model to estimate latent environments. Additionally, we further devise the modular prior networks to disentangle stationary/nonstationary latent variables. These two modules realize automatic adaptation and enhance nonstationary forecasting performance. Experimental results on several datasets validate the effectiveness of our approach.

cs.LG↗

arXiv2025

When Selection Meets Intervention: Additional Complexities in Causal Discovery

We address the common yet often-overlooked selection bias in interventional studies, where subjects are selectively enrolled into experiments. For instance, participants in a drug trial are usually patients of the relevant disease; A/B tests on mobile applications target existing users only, and gene perturbation studies typically focus on specific cell types, such as cancer cells. Ignoring this bias leads to incorrect causal discovery results. Even when recognized, the existing paradigm for interventional causal discovery still fails to address it. This is because subtle differences in when and where interventions happen can lead to significantly different statistical patterns. We capture this dynamic by introducing a graphical model that explicitly accounts for both the observed world (where interventions are applied) and the counterfactual world (where selection occurs while interventions have not been applied). We characterize the Markov property of the model, and propose a provably sound algorithm to identify causal relations as well as selection mechanisms up to the equivalence class, from data with soft interventions and unknown targets. Through synthetic and real-world experiments, we demonstrate that our algorithm effectively identifies true causal relations despite the presence of selection bias.

cs.LG↗

arXiv2025

A Model Fusion Distributed Kalman Filter For Non-Gaussian Observation Noise

Wireless sensor networks (WSNs) represent a critical research domain within the Internet of Things (IoT) technology. The distributed Kalman filter (DKF) has garnered significant attention as an information fusion method for WSNs. However, effectively handling non-Gaussian environments remains a crucial challenge for DKF. This paper proposes a solution by partitioning the noise distribution into multiple Gaussian components, thereby approximating the measurement model with sub-models. We introduce a model fusion distributed Kalman filter (MFDKF) that combines sub-models by assuming independent random processes for the model's transition probabilities. The expectation maximization (EM) algorithm is employed to estimate the relevant parameters. To address specific requirements in WSNs that demand high consensus or have limited communication, two derivative algorithms, namely consensus MFDKF (C-MFDKF) and simplified MFDKF (S-MFDKF), are proposed based on consensus theory. The convergence of MFDKF and its derivative algorithms is analyzed. A series of simulations demonstrate the effectiveness of MFDKF and its derivative algorithms. This paper has been accepted and published in IEEE Internet of Things Journal (Early Access), DOI: https://doi.org/10.1109/JIOT.2025.3526240

eess.SP↗

arXiv2025

Infant Cry Detection Using Causal Temporal Representation

This paper addresses a major challenge in acoustic event detection, in particular infant cry detection in the presence of other sounds and background noises: the lack of precise annotated data. We present two contributions for supervised and unsupervised infant cry detection. The first is an annotated dataset for cry segmentation, which enables supervised models to achieve state-of-the-art performance. Additionally, we propose a novel unsupervised method, Causal Representation Spare Transition Clustering (CRSTC), based on causal temporal representation, which helps address the issue of data scarcity more generally. By integrating the detected cry segments, we significantly improve the performance of downstream infant cry classification, highlighting the potential of this approach for infant care applications.

cs.SD↗

arXiv2025

Can Large Language Models Help Experimental Design for Causal Discovery?

Designing proper experiments and selecting optimal intervention targets is a longstanding problem in scientific or causal discovery. Identifying the underlying causal structure from observational data alone is inherently difficult. Obtaining interventional data, on the other hand, is crucial to causal discovery, yet it is usually expensive and time-consuming to gather sufficient interventional data to facilitate causal discovery. Previous approaches commonly utilize uncertainty or gradient signals to determine the intervention targets. However, numerical-based approaches may yield suboptimal results due to the inaccurate estimation of the guiding signals at the beginning when with limited interventional data. In this work, we investigate a different approach, whether we can leverage Large Language Models (LLMs) to assist with the intervention targeting in causal discovery by making use of the rich world knowledge about the experimental design in LLMs. Specifically, we present Large Language Model Guided Intervention Targeting (LeGIT) -- a robust framework that effectively incorporates LLMs to augment existing numerical approaches for the intervention targeting in causal discovery. Across 4 realistic benchmark scales, LeGIT demonstrates significant improvements and robustness over existing methods and even surpasses humans, which demonstrates the usefulness of LLMs in assisting with experimental design for scientific discovery.

cs.AI↗

arXiv2025

Quantum Charging Advantage from Multipartite Entanglement

Collective quantum batteries (QBs) demonstrate remarkable acceleration in charging dynamics compared to their individual counterparts, underscoring the pivotal contribution of quantum correlations to advanced energy storage paradigms. A fundamental challenge lies in identifying QBs that exhibit genuine quantum advantages derived from multipartite entanglement. In this Letter, based on numerical and analytical evidence, we conjecture a universal bound on the charging rate for fully charging schemes, which is determined by the maximum entanglement depth arising during the charging dynamics. Here, the charging rate quantifies the intrinsic evolution speed of the charging process, appropriately normalized against the quantum speed limit (QSL). We analytically validate this conjecture in three distinct scenarios: (i) fully charging schemes saturating the QSL, (ii) fully parallel charging schemes, and (iii) the SU(2) fully charging schemes. Moreover, we establish a novel lower bound for entanglement depth detection, facilitating numerical verification of our proposed conjecture. By defining the genuine quantum charging advantage as the ratio between entanglement-enhanced charging rates and the maximum achievable non-entangling charging rate, we demonstrate that the charging rate constitutes a robust indicator of genuine quantum advantages.

quant-ph↗

arXiv2025

Prompting Fairness: Integrating Causality to Debias Large Language Models

Large language models (LLMs), despite their remarkable capabilities, are susceptible to generating biased and discriminatory responses. As LLMs increasingly influence high-stakes decision-making (e.g., hiring and healthcare), mitigating these biases becomes critical. In this work, we propose a causality-guided debiasing framework to tackle social biases, aiming to reduce the objectionable dependence between LLMs' decisions and the social information in the input. Our framework introduces a novel perspective to identify how social information can affect an LLM's decision through different causal pathways. Leveraging these causal insights, we outline principled prompting strategies that regulate these pathways through selection mechanisms. This framework not only unifies existing prompting-based debiasing techniques, but also opens up new directions for reducing bias by encouraging the model to prioritize fact-based reasoning over reliance on biased social cues. We validate our framework through extensive experiments on real-world datasets across multiple domains, demonstrating its effectiveness in debiasing LLM decisions, even with only black-box access to the model.

cs.CL↗

arXiv2025

Synergy Between Sufficient Changes and Sparse Mixing Procedure for Disentangled Representation Learning

Disentangled representation learning aims to uncover latent variables underlying the observed data, and generally speaking, rather strong assumptions are needed to ensure identifiability. Some approaches rely on sufficient changes on the distribution of latent variables indicated by auxiliary variables such as domain indices, but acquiring enough domains is often challenging. Alternative approaches exploit structural sparsity assumptions on the mixing procedure, but such constraints are usually (partially) violated in practice. Interestingly, we find that these two seemingly unrelated assumptions can actually complement each other to achieve identifiability. Specifically, when conditioned on auxiliary variables, the sparse mixing procedure assumption provides structural constraints on the mapping from estimated to true latent variables and hence compensates for potentially insufficient distribution changes. Building on this insight, we propose an identifiability theory with less restrictive constraints regarding distribution changes and the sparse mixing procedure, enhancing applicability to real-world scenarios. Additionally, we develop an estimation framework incorporating a domain encoding network and a sparse mixing constraint and provide two implementations based on variational autoencoders and generative adversarial networks, respectively. Experiment results on synthetic and real-world datasets support our theoretical results.

cs.LG↗

arXiv2025

Enhanced Derivative-Free Optimization Using Adaptive Correlation-Induced Finite Difference Estimators

Gradient-based methods are well-suited for derivative-free optimization (DFO), where finite-difference (FD) estimates are commonly used as gradient surrogates. Traditional stochastic approximation methods, such as Kiefer-Wolfowitz (KW) and simultaneous perturbation stochastic approximation (SPSA), typically utilize only two samples per iteration, resulting in imprecise gradient estimates and necessitating diminishing step sizes for convergence. In this paper, we first explore an efficient FD estimate, referred to as correlation-induced FD estimate, which is a batch-based estimate. Then, we propose an adaptive sampling strategy that dynamically determines the batch size at each iteration. By combining these two components, we develop an algorithm designed to enhance DFO in terms of both gradient estimation efficiency and sample efficiency. Furthermore, we establish the consistency of our proposed algorithm and demonstrate that, despite using a batch of samples per iteration, it achieves the same convergence rate as the KW and SPSA methods. Additionally, we propose a novel stochastic line search technique to adaptively tune the step size in practice. Finally, comprehensive numerical experiments confirm the superior empirical performance of the proposed algorithm.

math.OC↗

arXiv2025

Colloidal Magnus effect in polymer solutions

Rotating particles moving in fluids undergo a transverse migration via the inertia-induced Magnus effect. This phenomenon vanishes at colloidal scales because inertia is negligible and the fluid flow is time reversible. Yet, recent experiments discovered an inverse Magnus effect of colloids in polymeric and micellar solutions supposedly because their viscoelasticity breaks the time reversibility. Our study shows that classical viscoelastic features -- normal-stress differences and/or shear-thinning cannot explain this phenomenon. Instead, it originates from local polymer density inhomogeneities due to their stress-gradient-induced transport, a mechanism increasingly important at smaller scales -- indeed relevant to colloidal experiments. Incorporating this mechanism into our model leads to quantitative agreement with the experiments without fitting parameters. Our work provides new insights into colloidal motion in complex fluids with microstructural inhomogeneities, offers a simple mechanistic theory for predicting the resulting migration, and underscores the necessity of assimilating these findings in future designs of micro-machinery including swimmers, actuators, rheometers, and so on.

cond-mat.soft↗

arXiv2025

Flow: Modularized Agentic Workflow Automation

Multi-agent frameworks powered by large language models (LLMs) have demonstrated great success in automated planning and task execution. However, the effective adjustment of agentic workflows during execution has not been well studied. An effective workflow adjustment is crucial in real-world scenarios, as the initial plan must adjust to unforeseen challenges and changing conditions in real time to ensure the efficient execution of complex tasks. In this paper, we define workflows as an activity-on-vertex (AOV) graph, which allows continuous workflow refinement by LLM agents through dynamic subtask allocation adjustment based on historical performance and previous AOVs. To further enhance framework performance, we emphasize modularity in workflow design based on evaluating parallelism and dependency complexity. With this design, our proposed multi-agent framework achieves efficient concurrent execution of subtasks, effective goal achievement, and enhanced error tolerance. Empirical results across various practical tasks demonstrate significant improvements in the efficiency of multi-agent frameworks through dynamic workflow refinement and modularization. The code is available at: https://github.com/tmllab/2025_ICLR_FLOW.

cs.AI↗

arXiv2025

Entanglement and energy transportation in the central-spin quantum battery

Quantum battery exploits the principle of quantum mechanics to transport and store energy. We study the energy transportation of the central-spin quantum battery, which is composed of $N_b$ spins serving as the battery cells, and surrounded by $N_c$ spins serving as the charger cells. We apply the invariant subspace method to solve the dynamics of the central-spin battery with a large number of spins. We establish a universal inverse relationship between the battery capacity and the battery-charger entanglement, which persists in any size of the battery and charger cells. Moreover, we find that when $N_b=N_c$, the central-spin battery has the optimal energy transportation, corresponding to the minimal battery-charger entanglement. Surprisingly, the central-spin battery has a uniform energy transportation behaviors in certain battery-charger scales. Our results reveal a nonmonotonic relationship between the battery-charger size and the energy transportation efficiency, which may provide more insights on designing other types of quantum batteries.

quant-ph↗

arXiv2025

On the Parameter Identifiability of Partially Observed Linear Causal Models

Linear causal models are important tools for modeling causal dependencies and yet in practice, only a subset of the variables can be observed. In this paper, we examine the parameter identifiability of these models by investigating whether the edge coefficients can be recovered given the causal structure and partially observed data. Our setting is more general than that of prior research - we allow all variables, including both observed and latent ones, to be flexibly related, and we consider the coefficients of all edges, whereas most existing works focus only on the edges between observed variables. Theoretically, we identify three types of indeterminacy for the parameters in partially observed linear causal models. We then provide graphical conditions that are sufficient for all parameters to be identifiable and show that some of them are provably necessary. Methodologically, we propose a novel likelihood-based parameter estimation method that addresses the variance indeterminacy of latent variables in a specific way and can asymptotically recover the underlying parameters up to trivial indeterminacy. Empirical studies on both synthetic and real-world datasets validate our identifiability theory and the effectiveness of the proposed method in the finite-sample regime. Code: https://github.com/dongxinshuai/scm-identify.

cs.LG↗

arXiv2025

Optimal energy storage and collective charging speedup in the central-spin quantum battery

Quantum batteries (QBs) exploit principles of quantum mechanics to accelerate the charging process and aim to achieve optimal energy storage. However, analytical results for investigating these problems remain lacking due to the challenges associated with nonequilibrium dynamics. In this work, we analytically investigate a central-spin QB model in which $N_b$ spin-1/2 battery cells interact with $N_c$ spin-1/2 charger units, using $m$ initially excited charger units as a resource. By employing the invariant subspace method and the shifted Holstein-Primakoff transformation, we identify four scenarios in which optimal energy storage can be achieved: (i) $N_b\!\ll\!m\!\ll\!N_c$; (ii) $m\!\ll\!N_b\!\ll\!N_c$; (iii) $m\!\ll\!N_c\!\ll\!N_b$; and (iv) $N_b\!\ll\!m\!=\!kN_c$ [$k\!\in\!(0,1)$]. In these cases, optimal storage is ensured by the SU(2) symmetry emerging from the charging dynamics. The first three cases map the central-spin QB to different Tavis-Cummings (TC) QBs, while the fourth corresponds to the non-TC limit. We analytically determine the charging time and demonstrate that in the fully charging cases (i) and (iv), the collective charging exhibits an $N_b$-fold enhancement in speedup compared to the parallel charging scheme. Additionally, we numerically observe a unified charging behavior when $m\!=\!N_c$, showing that asymptotically optimal energy storage is possible when $N_b\!=\!m\!=\!N_c$. In this case, we find a collective charging enhancement scaling as $N_b^{0.8264}$. The origin of the collective charging advantage in central-spin quantum batteries is also analyzed through the quantum speed limit and a multipartite entanglement witness. Our results highlight the crucial role of dynamically emergent SU(2) symmetry in providing an analytical understanding of non-equilibrium charging dynamics in QBs.

quant-ph↗