arXiv subjects

Kexin Liu

Publications and source records attributed to Kexin Liu.

At least 19 recordsLinked to original sources

arXiv2026

A visual large language foundational model for medical image recognition using clinician-oriented social media

Large language models (LLMs) have demonstrated strong capabilities across diverse domains, showing considerable potential in medicine. However, their application in medical settings remains limited by the scarcity of visual question answering (VQA) datasets that capture clinical reasoning and explicit image-text alignment. Here, we leverage de-identified medical images and expert commentaries shared on clinician-oriented social media. By combining an advanced LLM with clinician-in-the-loop verification, we established a rigorous pipeline to construct ThoughtMed-1M, a long-form medical VQA dataset containing over one million VQA pairs and designed to capture structured clinical logic and medical image-text alignment. To demonstrate its utility, we developed a FOundational LLM Trained on ThoughtMed-1M (FOLTMed). FOLTMed achieved state-of-the-art performance across 42 medical VQA benchmark datasets, with a macro accuracy of 85.4%, and generated more clinically coherent responses on the ThoughtMed-1M test set. It outperformed state-of-the-art models by 3--5% across factuality and similarity metrics, highlighting a scalable paradigm for advancing research on clinically grounded multimodal LLMs.

cs.AI↗

arXiv2026

The VariableTNG project: mass-dependent regulation of galaxy morphology by baryonic feedback

Galaxy morphology is shaped by both assembly history and baryonic processes, but their relative roles remain uncertain. We use the VariableTNG (VTNG) simulation suite to investigate how variations in baryonic feedback regulate galaxy morphology while keeping the initial conditions fixed. VTNG consists of cosmological magnetohydrodynamic simulations performed with the moving-mesh code {\sc AREPO}, varying eight parameters governing stellar and AGN feedback. At $z=1$, we characterize morphology using $\kappa_{\rm co}$, $v/\sigma$, and the axis ratio $c/a$. We find substantial morphological diversity across feedback models, with the dominant mechanism strongly dependent on stellar mass. For $M_\ast \lesssim 10^{11}\,{\rm M_\odot}$, morphology is primarily controlled by the supernova temperature $T_{\rm SN}$: larger $T_{\rm SN}$ delays early star formation, promotes a denser and more rotationally supported gas reservoir, and favours subsequent disc growth through in-situ star formation. At higher masses, morphology becomes increasingly sensitive to AGN feedback, particularly the quasar-mode coupling efficiency $\epsilon_{\rm f,high}$. Higher $\epsilon_{\rm f,high}$ suppresses early black hole growth, thereby weakening subsequent radio-mode feedback associated with gas depletion and loss of rotational support. The resulting morphology--mass relation is non-monotonic, with maximum rotational support at intermediate stellar masses. Our results demonstrate that baryonic feedback regulates galaxy morphology through distinct mass-dependent pathways, with stellar feedback dominating at lower masses and AGN self-regulation becoming increasingly important at the massive end.

astro-ph.GA↗

arXiv2026

The VariableTNG project: how baryonic mechanisms shape galaxy properties

We use 50 Sobol-sampled VariableTNG simulations, varying eight subgrid parameters at fixed cosmology and initial conditions, to determine which baryonic processes regulate galaxies and black holes at $z=6$ and $z=8$. We compare the simulations with recent high-redshift measurements of the stellar mass function, star-forming main sequence, stellar and gas-phase mass-metallicity relations, stellar mass-size relation, and black-hole host and accretion properties. The simulations reproduce several broad trends in these observables, although differences remain in gas-phase metallicity, galaxy size, and the most extreme black-hole populations. Given the substantial uncertainties in both the physical modelling and the observational inference of high-redshift galaxy properties, we regard these differences as diagnostic tensions rather than definitive model failures. Random-forest analyses reveal a clear hierarchy in parameter sensitivity. The abundance of low-mass galaxies is regulated primarily by stellar feedback, particularly the supernova temperature $T_{\mathrm{SN}}$ and thermal wind fraction $\tau_{\mathrm{w}}$, whereas the sensitivity of the scatter in the star-forming main sequence is weaker and redshift dependent. The high-accretion tail of black-hole growth depends on black-hole seeding and feedback parameters, but also on $T_{\mathrm{SN}}$, suggesting that stellar feedback indirectly regulates rapid black-hole growth through its impact on the available gas supply. Although cosmic variance can obscure these intrinsic responses in independent small volumes, our controlled experiment identifies stellar feedback as a common physical link between early low-mass galaxy formation and rapid black-hole growth.

astro-ph.GA↗

arXiv2026

The VariableTNG project: Unveiling the physical drivers of galaxy quenching

Understanding the physical processes that regulate galaxy quenching is a key challenge in galaxy formation and evolution. The VariableTNG (VTNG) project provides a laboratory to investigate these processes, as it systematically varies eight parameters of galaxy formation while keeping the initial conditions fixed, allowing the effects of individual feedback prescriptions to be isolated. We use interpretable machine-learning techniques as a tool to identify the parameters that most strongly regulate the quenched galaxy fraction. Our goal is to quantify the relative importance of the galaxy formation and feedback parameters that regulate the quenched galaxy fraction at z=0, determine how their influence changes across stellar mass and environment. We compute the quenched galaxy fraction for 26 VTNG boxes as a function of stellar mass, black hole mass, and gas mass, considering both the total galaxy population and separate samples of central and satellite galaxies. We train Random Forest regressors to predict the variation of the quenched fraction relative to the TNG100-1 model and use SHAP values to quantify both the magnitude and direction of the influence of each parameter. Our analysis reveals that only a small subset of the VTNG parameters dominates the variance of the quenched fraction. The stellar feedback wind parameter is the primary driver at low stellar masses, while its importance gradually shifts toward AGN-related parameters at higher masses. The supernova temperature also plays an important role at both extremes of the stellar-mass range. This transition persists when the galaxy population is divided into central and satellite systems. Comparisons with observational measurements further suggest that variations in these feedback parameters may contribute to the discrepancies between the fiducial TNG100-1 model and the observed passive galaxy population.

astro-ph.GA↗

arXiv2026

Learning domain-invariant features through channel-level sparsification for Out-Of Distribution Generalization

Out-of-Distribution (OOD) generalization has become a primary metric for evaluating image analysis systems. Since deep learning models tend to capture domain-specific context, they often develop shortcut dependencies on these non-causal features, leading to inconsistent performance across different data sources. Current techniques, such as invariance learning, attempt to mitigate this. However, they struggle to isolate highly mixed features within deep latent spaces. This limitation prevents them from fully resolving the shortcut learning problem.In this paper, we propose Hierarchical Causal Dropout (HCD), a method that uses channel-level causal masks to enforce feature sparsity. This approach allows the model to separate causal features from spurious ones, effectively performing a causal intervention at the representation level. The training is guided by a Matrix-based Mutual Information (MMI) objective to minimize the mutual information between latent features and domain labels, while simultaneously maximizing the information shared with class labels.To ensure stability, we incorporate a StyleMix-driven VICReg module, which prevents the masks from accidentally filtering out essential causal data. Experimental results on OOD benchmarks show that HCD performs better than existing top-tier methods.

cs.CV↗

arXiv2026

Integrated cooperative localization of heterogeneous measurement swarm: A unified data-driven method

The cooperative localization (CL) problem in heterogeneous robotic systems with different measurement capabilities is investigated in this work. In practice, heterogeneous sensors lead to directed and sparse measurement topologies, whereas most existing CL approaches rely on multilateral localization with restrictive multi-neighbor geometric requirements. To overcome this limitation, we enable pairwise relative localization (RL) between neighboring robots using only mutual measurement and odometry information. A unified data-driven adaptive RL estimator is first developed to handle heterogeneous and unidirectional measurements. Based on the convergent RL estimates, a distributed pose-coupling CL strategy is then designed, which guarantees CL under a weakly connected directed measurement topology, representing the least restrictive condition among existing results. The proposed method is independent of specific control tasks and is validated through a formation control application and real-world experiments.

cs.RO↗

arXiv2025

Whole-Body Control With Terrain Estimation of A 6-DoF Wheeled Bipedal Robot

Wheeled bipedal robots have garnered increasing attention in exploration and inspection. However, most research simplifies calculations by ignoring leg dynamics, thereby restricting the robot's full motion potential. Additionally, robots face challenges when traversing uneven terrain. To address the aforementioned issue, we develop a complete dynamics model and design a whole-body control framework with terrain estimation for a novel 6 degrees of freedom wheeled bipedal robot. This model incorporates the closed-loop dynamics of the robot and a ground contact model based on the estimated ground normal vector. We use a LiDAR inertial odometry framework and improved Principal Component Analysis for terrain estimation. Task controllers, including PD control law and LQR, are employed for pose control and centroidal dynamics-based balance control, respectively. Furthermore, a hierarchical optimization approach is used to solve the whole-body control problem. We validate the performance of the terrain estimation algorithm and demonstrate the algorithm's robustness and ability to traverse uneven terrain through both simulation and real-world experiments.

cs.RO↗

arXiv2025

Mean-Shift Theory and Its Applications in Swarm Robotics: A New Way to Enhance the Efficiency of Multi-Robot Collaboration

Swarms evolving from collective behaviors among multiple individuals are commonly seen in nature, which enables biological systems to exhibit more efficient and robust collaboration. Creating similar swarm intelligence in engineered robots poses challenges to the design of collaborative algorithms that can be programmed at large scales. The assignment-based method has played an eminent role for a very long time in solving collaboration problems of robot swarms. However, it faces fundamental limitations in terms of efficiency and robustness due to its unscalability to swarm variants. This article presents a tutorial review on recent advances in assignment-free collaboration of robot swarms, focusing on the problem of shape formation. A key theoretical component is the recently developed \emph{mean-shift exploration} strategy, which improves the collaboration efficiency of large-scale swarms by dozens of times. Further, the efficiency improvement is more significant as the swarm scale increases. Finally, this article discusses three important applications of the mean-shift exploration strategy, including precise shape formation, area coverage formation, and maneuvering formation, as well as their corresponding industrial scenarios in smart warehousing, area exploration, and cargo transportation.

cs.RO↗

arXiv2025

Causally Guided Gaussian Perturbations for Out-Of-Distribution Generalization in Medical Imaging

Out-of-distribution (OOD) generalization remains a central challenge in deploying deep learning models to real-world scenarios, particularly in domains such as biomedical images, where distribution shifts are both subtle and pervasive. While existing methods often pursue domain invariance through complex generative models or adversarial training, these approaches may overlook the underlying causal mechanisms of generalization.In this work, we propose Causally-Guided Gaussian Perturbations (CGP)-a lightweight framework that enhances OOD generalization by injecting spatially varying noise into input images, guided by soft causal masks derived from Vision Transformers. By applying stronger perturbations to background regions and weaker ones to foreground areas, CGP encourages the model to rely on causally relevant features rather than spurious correlations.Experimental results on the challenging WILDS benchmark Camelyon17 demonstrate consistent performance gains over state-of-the-art OOD baselines, highlighting the potential of causal perturbation as a tool for reliable and interpretable generalization.

cs.CV↗

arXiv2025

Bidirectional Task-Motion Planning Based on Hierarchical Reinforcement Learning for Strategic Confrontation

In swarm robotics, confrontation scenarios, including strategic confrontations, require efficient decision-making that integrates discrete commands and continuous actions. Traditional task and motion planning methods separate decision-making into two layers, but their unidirectional structure fails to capture the interdependence between these layers, limiting adaptability in dynamic environments. Here, we propose a novel bidirectional approach based on hierarchical reinforcement learning, enabling dynamic interaction between the layers. This method effectively maps commands to task allocation and actions to path planning, while leveraging cross-training techniques to enhance learning across the hierarchical framework. Furthermore, we introduce a trajectory prediction model that bridges abstract task representations with actionable planning goals. In our experiments, it achieves over 80% in confrontation win rate and under 0.01 seconds in decision time, outperforming existing approaches. Demonstrations through large-scale tests and real-world robot experiments further emphasize the generalization capabilities and practical applicability of our method.

cs.RO↗

arXiv2024

Relative Pose Estimation for Nonholonomic Robot Formation with UWB-IO Measurements (Extended version)

This article studies the problem of distributed formation control for multiple robots by using onboard ultra wide band (UWB) distance and inertial odometer (IO) measurements. Although this problem has been widely studied, a fundamental limitation of most works is that they require each robot's pose and sensor measurements are expressed in a common reference frame. However, it is inapplicable for nonholonomic robot formations due to the practical difficulty of aligning IO measurements of individual robot in a common frame. To address this problem, firstly, a concurrent-learning based estimator is firstly proposed to achieve relative localization between neighboring robots in a local frame. Different from most relative localization methods in a global frame, both relative position and orientation in a local frame are estimated with only UWB ranging and IO measurements. Secondly, to deal with information loss caused by directed communication topology, a cooperative localization algorithm is introduced to estimate the relative pose to the leader robot. Thirdly, based on the theoretical results on relative pose estimation, a distributed formation tracking controller is proposed for nonholonomic robots. Both 3D and 2D real-world experiments conducted on aerial robots and grounded robots are provided to demonstrate the effectiveness of the proposed method.

cs.RO↗

arXiv2024

Distributed Formation Shape Control of Identity-less Robot Swarms

Different from most of the formation strategies where robots require unique labels to identify topological neighbors to satisfy the predefined shape constraints, we here study the problem of identity-less distributed shape formation in homogeneous swarms, which is rarely studied in the literature. The absence of identities creates a unique challenge: how to design appropriate target formations and local behaviors that are suitable for identity-less formation shape control. To address this challenge, we propose the following novel results. First, to avoid using unique identities, we propose a dynamic formation description method and solve the formation consensus of robots in a locally distributed manner. Second, to handle identity-less distributed formations, we propose a fully distributed control law for homogeneous swarms based on locally sensed information. While the existing methods are applicable to simple cases where the target formation is stationary, ours can tackle more general maneuvering formations such as translation, rotation, or even shape deformation. Both numerical simulation and flight experiment are presented to verify the effectiveness and robustness of our proposed formation strategy.

cs.RO↗

arXiv2024

Concurrent-Learning Based Relative Localization in Shape Formation of Robot Swarms (Extended version)

In this paper, we address the shape formation problem for massive robot swarms in environments where external localization systems are unavailable. Achieving this task effectively with solely onboard measurements is still scarcely explored and faces some practical challenges. To solve this challenging problem, we propose the following novel results. Firstly, to estimate the relative positions among neighboring robots, a concurrent-learning based estimator is proposed. It relaxes the persistent excitation condition required in the classical ones such as least-square estimator. Secondly, we introduce a finite-time agreement protocol to determine the shape location. This is achieved by estimating the relative position between each robot and a randomly assigned seed robot. The initial position of the seed one marks the shape location. Thirdly, based on the theoretical results of the relative localization, a novel behavior-based control strategy is devised. This strategy not only enables adaptive shape formation of large group of robots but also enhances the observability of inter-robot relative localization. Numerical simulation results are provided to verify the performance of our proposed strategy compared to the state-of-the-art ones. Additionally, outdoor experiments on real robots further demonstrate the practical effectiveness and robustness of our methods.

cs.RO↗

arXiv2024

Disparate Effects of Circumgalactic Medium Angular Momentum in IllustrisTNG and SIMBA

In this study, we examine the role of circumgalactic medium (CGM) angular momentum ($j_{\rm CGM}$) on star formation in galaxies, whose influence is currently not well understood. The analysis utilises central galaxies from two hydrodynamical simulations, SIMBA and IllustrisTNG. We observe a substantial divergence in how star formation rates correlate with CGM angular momentum between the two simulations. Specifically, quenched galaxies in IllustrisTNG show high $j_{\rm CGM}$, while in SIMBA, quenched galaxies have low $j_{\rm CGM}$. This difference is attributed to the distinct active galactic nucleus (AGN) feedback mechanisms active in each simulation. Moreover, both simulations demonstrate similar correlations between $j_{\rm CGM}$ and environmental angular momentum ($j_{\rm Env}$) in star-forming galaxies, but these correlations change notably when kinetic AGN feedback is present. In IllustrisTNG, quenched galaxies consistently show higher $j_{\rm CGM}$ compared to their star-forming counterparts with the same $j_{\rm Env}$, a trend not seen in SIMBA. Examining different AGN feedback models in SIMBA, we further confirm that AGN feedback significantly influences the CGM gas distribution, although the relationship between the cold gas fraction and the star formation rate (SFR) remains largely stable across different feedback scenarios.

astro-ph.GA↗

arXiv2024

Scalable DAQ system operating the CHIPS-5 neutrino detector

The CHIPS R&D project focuses on development of low-cost water Cherenkov neutrino detectors through novel design strategies and resourceful engineering. This work presents an end-to-end DAQ solution intended for a recent 5 kt CHIPS prototype, which is largely based on affordable mass-produced components. Much like the detector itself, the presented instrumentation is composed of modular arrays that can be scaled up and easily serviced. A single such array can carry up to 30 photomultiplier tubes (PMTs) accompanied by electronics that generate high voltage in-situ and deliver time resolution of up to 0.69 ns. In addition, the technology is compatible with the White Rabbit timing system, which can synchronize its elements to within 100 ps. While deployment issues did not permit the presented DAQ system to operate beyond initial evaluation, the presented hardware and software successfully passed numerous commissioning tests that demonstrated their viability for use in a large-scale neutrino detector, instrumented with thousands of PMTs.

physics.ins-det↗

arXiv2024

Hierarchical Reinforcement Learning for Swarm Confrontation with High Uncertainty

In swarm robotics, confrontation including the pursuit-evasion game is a key scenario. High uncertainty caused by unknown opponents' strategies, dynamic obstacles, and insufficient training complicates the action space into a hybrid decision process. Although the deep reinforcement learning method is significant for swarm confrontation since it can handle various sizes, as an end-to-end implementation, it cannot deal with the hybrid process. Here, we propose a novel hierarchical reinforcement learning approach consisting of a target allocation layer, a path planning layer, and the underlying dynamic interaction mechanism between the two layers, which indicates the quantified uncertainty. It decouples the hybrid process into discrete allocation and continuous planning layers, with a probabilistic ensemble model to quantify the uncertainty and regulate the interaction frequency adaptively. Furthermore, to overcome the unstable training process introduced by the two layers, we design an integration training method including pre-training and cross-training, which enhances the training efficiency and stability. Experiment results in both comparison, ablation, and real-robot studies validate the effectiveness and generalization performance of our proposed approach. In our defined experiments with twenty to forty agents, the win rate of the proposed method reaches around ninety percent, outperforming other traditional methods.

cs.RO↗

arXiv2023

Multi-Agent Reinforcement Learning-Based UAV Pathfinding for Obstacle Avoidance in Stochastic Environment

Traditional methods plan feasible paths for multiple agents in the stochastic environment. However, the methods' iterations with the changes in the environment result in computation complexities, especially for the decentralized agents without a centralized planner. Although reinforcement learning provides a plausible solution because of the generalization for different environments, it struggles with enormous agent-environment interactions in training. Here, we propose a novel centralized training with decentralized execution method based on multi-agent reinforcement learning, which is improved based on the idea of model predictive control. In our approach, agents communicate only with the centralized planner to make decentralized decisions online in the stochastic environment. Furthermore, considering the communication constraint with the centralized planner, each agent plans feasible paths through the extended observation, which combines information on neighboring agents based on the distance-weighted mean field approach. Inspired by the rolling optimization approach of model predictive control, we conduct multi-step value convergence in multi-agent reinforcement learning to enhance the training efficiency, which reduces the expensive interactions in convergence. Experiment results in both comparison, ablation, and real-robot studies validate the effectiveness and generalization performance of our method.

cs.RO↗

arXiv2023

Model predictive control-based value estimation for efficient reinforcement learning

Reinforcement learning suffers from limitations in real practices primarily due to the number of required interactions with virtual environments. It results in a challenging problem because we are implausible to obtain a local optimal strategy with only a few attempts for many learning methods. Hereby, we design an improved reinforcement learning method based on model predictive control that models the environment through a data-driven approach. Based on the learned environment model, it performs multi-step prediction to estimate the value function and optimize the policy. The method demonstrates higher learning efficiency, faster convergent speed of strategies tending to the local optimal value, and less sample capacity space required by experience replay buffers. Experimental results, both in classic databases and in a dynamic obstacle avoidance scenario for an unmanned aerial vehicle, validate the proposed approaches.

cs.LG↗