arXiv subjects

He Wang

Publications and source records attributed to He Wang.

At least 19 recordsLinked to original sources

arXiv2026

ReflexArc: Bottom-Up Integration of Instantaneous Tactile Reflexes into Whole-Body Humanoid Control

Humanoid robots operating in cluttered environments must react to unexpected contacts before the first impact peak, which arrives within tens of milliseconds. Learned whole-body policies running at 50 Hz introduce a policy-cycle delay and may suppress a protective withdrawal as a tracking error. To address these challenges, we propose ARC (Ascending Reflex Control), a bottom-up framework that integrates instantaneous tactile reflexes with learned whole-body control. ARC consists of two components: an instantaneous reflex pathway and bottom-up reflex integration. The reflex pathway detects hazardous contacts from full-body tactile sensing, solves a whole-body withdrawal pose, and delivers it directly to the motors within 10 ms. Bottom-up reflex integration combines reflex-in-the-loop policy learning, reflex-transparent policy alignment, and reflex-informed policy adaptation to sustain the withdrawal, maintain balance, and complete longer-horizon recovery. We deploy ARC on a Unitree G1 and evaluate three categories of hazardous contact. Compared with an otherwise identical reflex delayed by 20 ms, ARC reduces the first force peak by 34.1% and reduces the contact impulse from 0.2632 N s to 0.1225 N s (a 2.15x ratio). We further demonstrate protected motion tracking, teleoperation, and vision-free obstacle traversal guided by foot tactile sensing. Project page: https://minusmodulo.github.io/reflexarc-website/

cs.RO↗

arXiv2026

Large Language Model-Assisted Adaptive Heuristic Search for Agile Earth Observation Satellite Scheduling

Agile Earth observation satellites must coordinate imaging tasks within limited observation windows, while accounting for attitude maneuvering and onboard energy constraints. Competition among observation opportunities makes scheduling performance sensitive to the heuristic rules used for task selection. This letter proposes a large language model (LLM)-assisted adaptive heuristic search method for agile Earth observation satellite scheduling. The LLM generates an initial heuristic rule according to the characteristics of the scheduling instance. During the subsequent ant colony search, search performance and scheduling information are fed back to the LLM to refine the heuristic rule. The proposed modifications are evaluated through paired solution constructions before being incorporated into the search process. A unified scheduling procedure determines task selection, satellite assignment, and observation timing under the operational constraints. Experiments on six scheduling scenarios involving 50--200 imaging tasks demonstrate that the proposed method achieves 4.3\%--9.6\% higher mean observation benefit than the strongest baseline.

cs.AI↗

arXiv2026

Panoramic modelling of single-cell transcriptomes at scale with CellNet

Comprehensively characterizing cellular landscapes to elucidate cell diversity is a central goal in biology. Recent advances in single-cell profiling have created unprecedented opportunities for this endeavor by capturing diverse transcriptomes with remarkable breadth. However, current foundation models generally struggle to leverage such holistic and valuable information to represent a cell. Here we present CellNet, which enables panoramic modelling of single-cell transcriptomes at scale by integrating DNA-derived gene identities and measured RNA expression into a shared image representation. Pretrained on 231 million image-based cell profiles across 40 species and covering over one million genes, CellNet improves the recognition of rare and condition-associated cellular states, enables the study of genes beyond experimentally measured panels, and uncovers associations between poorly characterized non-coding RNAs and known biological programmes. CellNet paves the way for AI-enabled cell-biology discovery at the whole-transcriptome scale. The CellNet website is available at https://www.cellnet.top.

q-bio.GN↗

arXiv2026

Access Resources and Stability of Error-Correction Capabilities in Continuous Strategic Codes

Continuous quantum error correction depends on which parts of a measurement record remain available to the controller at recovery. We study this temporal access using continuous strategic codes, formulated as open continuous process tensors contracted with causal instruments and conditional recovery. For monitored stabilizer codes, a common-optimal-correction criterion identifies when records can be merged without loss and yields a factorization for separated CSS inference. Terminal decision labels and causal persistent memory carry different costs: five persistent classical states reach the complete-record optimum in a finite-strength counting model, whereas its two-window diffusive counterpart has nine terminal decision labels and requires unbounded memory for exact causal recovery. Uniform lower and constructive upper bounds establish an optimal fidelity deficit of order $m^{-2}$ for $m$ persistent states in the diffusive model. Counting and diffusive stability bounds translate open-process approximation into guarantees for deployed-controller performance and retained access advantages, quantified by explicit quantizers and model-mismatch curves. The same process--instrument framework connects storage with persistent environments and irreversible noise to spacetime interrogation and record-selected paths for a logical $T$ gate. These results specify which record distinctions recovery needs, the memory cost of approximating them, and the logical protection retained under model error.

quant-ph↗

arXiv2026

SAFE-SVD: Sensitivity-Aware Fidelity-Enforcing SVD for Physics Foundation Models

We propose a new method for compressing physics foundation models (PFMs) which is a new trend in AI for Science. While model compression is essential for reducing memory use and accelerating inference in large foundation models, it remains under-explored for PFMs, where preserving physical fidelity is crucial. The challenge lies in the functional nature of physics data, where partial derivatives encode spatiotemporal dynamics and exhibit high sensitivity to compression. Conventional compression methods ignore this structure, often causing severe performance degradation or failure. To address this, we introduce a sensitivity-aware fidelity-enforcing compression framework that explicitly models loss-aware layer sensitivity in the output function space during compression. This provides a new route to compressing scientific foundation models while preserving accuracy and physical fidelity. Experiments show substantial gains over existing methods across multiple models and datasets, achieving significantly higher compression ratios while maintaining accuracy, in some cases by orders of magnitude. More broadly, the work potentially leads to a new subfield of efficient, deployable, and sustainable scientific foundation models in AI for Science.

cs.LG↗

arXiv2026

Reinforcement Learning-Guided Evolutionary Policy Optimization for Preference-Adjustable Heterogeneous Agile Earth Observation Satellite Scheduling

Heterogeneous agile Earth observation satellite (AEOS) scheduling requires task selection, satellite assignment, and observation sequencing under satellite-dependent visibility windows, attitude maneuvering requirements, energy consumption, and onboard storage constraints. Since satellites differ in orbital access, maneuvering capability, and payload resources, the same task may have different feasible windows, transition costs, and resource-consumption patterns on different platforms, which increases the difficulty of unified modeling and efficient optimization. To address this problem, this paper proposes an evolutionary policy optimization framework for heterogeneous AEOS scheduling with preference-adjustable weighted objectives. In the modeling layer, assignment-based indirect encoding is combined with decoder-based equivalent-cost evaluation to retain satellite-dependent constraints while integrating task gain, energy saving, and load balance into an interpretable scalar utility. In the optimization layer, schedule decoding, population-based search, and online actor-critic operator control are decoupled, so that reinforcement learning selects high-level search operators rather than constructing schedules directly. Based on this framework, a reinforcement-learning-assisted operator-selection memetic evolutionary algorithm (RLOSMEA) is developed to coordinate global exploration, feasibility recovery, and local refinement under a limited function-evaluation budget. Experiments on different heterogeneous AEOS scenarios show that RLOSMEA achieves higher overall weighted utility and more stable convergence than representative metaheuristic baselines. Sensitivity and learning-behavior analyses further confirm the robustness of the proposed method and the effectiveness of reinforcement-learning-guided operator selection.

cs.AI↗

arXiv2026

Hard Constraint Projection in a Physics Informed Neural Network (Burgers)

Physics informed machine learning provides a framework to study and simulate fluid dynamics, potentially at a fraction of the computational cost of traditional numerical methods, with the goal of improving the explainability, accuracy and robustness of purely data driven machine learning methods. In this work, we extend the hard constraint projection (HCP) method developed by Chen et al. [Journal of Computational Physics, 445, 110624, 2021] from a linear partial differential equation (PDE) to a nonlinear PDE: the 1D Burgers' equation. To strictly impose a discrete approximation of the governing equation, the Burgers' equation is first locally discretised (using a finite difference stencil) then decomposed into a constraint matrix and a prediction matrix. Using tools from linear algebra, the constraint matrix is used to construct a projection matrix. A physics informed neural network (PINN) [Raissi et al. Journal of Computational Physics, 378, 686, 2019] is used to predict the velocity, and a final unlearnable layer projects the velocity prediction to the hyperplane of exact solutions to a discretised form of the Burgers' equation. Our initial results indicate that the HCP-PINN initially trains similarly to an equivalent PINN, and as the PINN loss plateaus, the HCP-PINN loss continues to decline. When tested on a previously unseen uniform grid of points covering the domain, the HCP-PINN exhibits higher accuracy, converging to a data error approximately an order of magnitude lower than the comparable PINN. We also observe that the inclusion of the HCP layer results in better observance of the governing PDE. Our findings suggest that the HCP-PINN can be more robust to the selection of physics collocation points, and less dependent on collocation point quantity.

physics.flu-dyn↗

arXiv2026

HumanoidNDM: Closing the Training-Deployment Gap of Motion Tracking via Humanoid Neural Dynamics Model

Humanoid robots have recently achieved remarkable success in motion tracking. However, during real-world deployment, robots inevitably encounter dynamics gaps arising from either extra physical disturbances or inaccurate simulation modeling, resulting in a significant training-deployment gap. Therefore, effectively leveraging real-world data to fine-tune policies represents the crucial last mile for robust deployment. To bridge this gap, we propose HumanoidNDM, an online fine-tuning framework for humanoid motion tracking driven by a real-world-trained neural dynamics model. First, in the data collection stage, we introduce a low-frequency reference perturbation method with an adaptive intensity mechanism, balancing deployment feasibility and exploration efficiency. Second, for the model training stage, we train a robust history-based, proprioceptive neural dynamics model in Euclidean space by incorporating differentiable forward kinematics, eliminating the reliance on costly motion capture systems. Finally, in the policy fine-tuning stage, we directly fine-tune the policy via backpropagated model gradients, achieving high data efficiency. Through physical validation on the Unitree G1 robot, we demonstrate that HumanoidNDM enables rapid real-world adaptation to various perturbations, representing the first online, MoCap-free fine-tuning framework for humanoid tracking policies.

cs.RO↗

arXiv2026

Reinforcement Learning for Adaptive Crowd Flow Control Using Dynamic Barriers

Crowd barriers are widely used to regulate pedestrian movement during large public events such as pilgrimages. However, fixed barrier layouts often create unnecessary stopping and local congestion even when they successfully control the overall crowd flow. Finding better barrier configurations is difficult because there is no dataset that provides the best barrier placement for a given crowd scene. We address this challenge by introducing a simulation-driven reinforcement learning framework that learns barrier configurations without requiring labelled training data. Starting from a segmented crowd scene, we construct a realistic simulation environment that reproduces the observed pedestrian distribution. This environment is then used to train a Soft Actor-Critic (SAC) agent that continuously adjusts barrier positions to improve pedestrian movement. Unlike supervised approaches the proposed framework allows the agent to explore barrier configurations that have never been observed in real data. We evaluate the learned policy under unseen numbers of pedestrians. The learned policy consistently outperforms the existing manually designed barrier configuration by reducing pedestrian stops while maintaining effective flow regulation. These results demonstrate that combining computer vision crowd simulation and reinforcement learning provides an effective framework for optimizing crowd control infrastructure.

cs.CV↗

arXiv2026

Tracking Is Not Sensing: Artifact-Aware Thermal Video Respiration Estimation in Head-Free Marmosets

Continuous physiological monitoring is essential for linking behavior, internal state, and neural activity in neuroscience and translational research. In common marmosets, a key nonhuman primate model for social neuroscience, respiratory rate provides a basic readout of arousal, stress, and physiological regulation. Contact sensors can provide accurate measurements, but they add preparation burden and may interfere with head-free behavioral protocols. Thermal video offers a noncontact alternative by capturing breathing-related temperature changes around the nose and mouth. However, in head-free marmosets, these cues are weak, localized, and easily corrupted by head motion, pose changes, and fur. Here we show that this setting exposes a broader challenge in video-based physiological sensing: visually reliable tracking does not necessarily imply physiologically valid measurement. Using synchronized thermal video and contact respiration recordings, we show that marmoset respiration is recoverable from thermal video but strongly ROI-sensitive. Under head-free motion, visually plausible nasal tracking can still produce traces inconsistent with contact respiration, demonstrating that tracking is not sensing. We therefore estimate respiratory rate from learned temporal signals extracted from thermal clips and introduce Selective Temporal Attenuation Regularization (STAR). STAR selectively attenuates high-motion thermal discontinuities through local temporal imputation during training, regularizing models toward respiration-consistent dynamics while leaving inference unchanged on raw video. Across multiple deep-learning backbones, STAR improves head-free estimation, reaching 2.23 bpm MAE compared with 20.76 bpm for tracked ROI analysis.

cs.CV↗

arXiv2026

WAM-TTT: Steering World-Action Models by Watching Human Play at Test Time

Steering robot foundation models (RFMs) toward new task variants or user-preferred behaviors remains challenging, often requiring additional robot demonstrations, task-specific fine-tuning, or long-context conditioning. We present WAM-TTT, a test-time training framework for steering world action models from raw human videos. Rather than treating human videos as trajectories to imitate, WAM-TTT absorbs them into a lightweight adaptive memory inside a frozen WAM through self-supervised video prediction. To make this memory useful for control, we introduce a meta-training stage that aligns human demonstrations with robot behaviors using paired human-robot data and a key--value memory reconstruction objective. At test time, only unlabeled human videos are required to adapt the memory, while the pretrained WAM remains frozen. This enables efficient and reusable steering without robot actions, human-side annotations, or task-specific fine-tuning, while preserving the generalization ability of the foundation model. Extensive experiments show that WAM-TTT consistently outperforms in-context human-video conditioning baselines across diverse manipulation tasks and generalization settings.

cs.RO↗

arXiv2026

Assessing the consistency of the CPL dark-energy model with cosmological data: Simulation-calibrated neural posterior disagreement

Recent DESI observations hint at a preference for dynamical dark energy with an equation of state that crosses the phantom divide within the Chevallier-Polarski-Linder (CPL) parameterization. In this work, we assess the consistency of the CPL parameterization with current cosmological observations. We develop a model check based on the disagreement between posteriors inferred by neural posterior estimation and flow-matching posterior estimation. Both estimators are trained on the same CPL simulations, and their posterior disagreement is calibrated using independent simulations to obtain a global $p$-value. Controlled simulations assess the reliability of the calibrated $p$-values and show how detection sensitivity depends on the form of departure from the assumed model. Applying the test to the compressed CMB+DESI+DESY5 data vector yields $p_{\rm global}=0.524$, indicating that the current observations are consistent with the CPL model. The method has a median evaluation time of only $0.53\,{\rm s}$, enabling efficient assessments of cosmological model adequacy. This framework enables efficient tests of dark-energy models, providing a firmer basis for investigating the physical origin of cosmic acceleration.

astro-ph.CO↗

arXiv2026

RLMuscle: Muscle-Driven Character Animation via Reinforcement Learning

Body deformation caused by muscle contraction plays a vital role in the visual realism of character animation, yet in most 3D animation pipelines artists have to manually specify muscle activations, which is time-consuming and gives no guarantee of physical realism. To address this, we propose a new method in which a reinforcement learning (RL) controller discovers per-muscle activations that track reference motions without manual tuning. Existing RL muscle controllers stop at 1D muscles; we go one step further, keeping the 1D model inside the RL control loop as a fast surrogate, and using a 3D volumetric simulation to produce the final deformation. Because both stages share the very same constitutive law, the muscle deformation is produced by the same activations that generated the motion. We further add a measured electromyography (EMG) prior that steers the redundant activation solution toward real human muscle recruitment. To achieve stable, efficient, and realistic simulation, we propose a novel geometric multigrid for Extended Position-Based Dynamics (XPBD) that is built on a coarse tetrahedral cage with barycentric-coordinate prolongation and closes a V-cycle on tetrahedral meshes. Within this GPU-parallelized solver, we inject the same Hill-type force-length law as an energy-based fiber constraint that turns per-fiber activation into realistic muscle contraction and bulging. We demonstrate volumetric muscle simulation on a 194-muscle human body with a diverse set of motions (walk, kick, swim, overhead-squat, etc.), simulating the full-body model at 18 ms per frame on a single RTX 4090. Our work takes a step toward character animation that is not only visually convincing but also biomechanically grounded. The website with code and videos is at https://chunleili.github.io/project-page-rlmuscle/.

cs.GR↗

arXiv2026

Where is Physical RSI capability from? An Observational Scaling Law of LLM-as-policy

In this paper, we systematically investigate the mechanisms, scaling behavior, and generalization of physical Recursive self-improvement. We introduce PhysRSI-eval to quantify physical RSI through a five-axis metric: First-Attempt Performance, Efficiency, Success, Budget and Improvement from failure. Different VLMs are evaluated on 18 manipulation tasks under a fixed interaction budget. The analysis of experimental results motivates the Physical Barrel Hypothesis: different capabilities may constrain different parts of improvement. The five-axis metrics also distinguish systems with close public benchmark averages and show that high scores need not come with early or efficient gains. We study observational scaling from two perspectives: predicting the relative performance of VLM agents from their public capabilities, and forecasting later progress from early physical feedback. Together, these analyses ask how quickly and how far an agent can improve. Moreover, we conduct workspace interventions and then analyze what improvement leaves behind. Removing history reduces late-round success, frozen workspaces remain useful on unseen layouts, and three cross-task cases achieve success earlier. The results suggest that scene-specific adaptation and reusable experience can coexist.

cs.RO↗

arXiv2026

Evolution and breakdown of chiral anomaly in Weyl semiconductor tellurium

Chiral anomaly in topological materials manifests as negative longitudinal magnetoresistance, yet the tunability and evolution of this effect remain elusive. Here we show that doped tellurium serves as a versatile platform to realize a giant negative longitudinal magnetoresistance that is highly tunable via the Fermi level and magnetic field. By moving the Fermi level towards the Weyl point, we observe a pronounced enhancement of negative magnetoresistance, which arises from chiral anomaly and robustly persists up to ambient temperature. Beyond the quantum limit, the system exhibits positive linear magnetoconductance, serving as a compelling transport signature of the chiral anomaly in the lowest Landau level. Notably, in the extreme quantum limit at cryogenic temperatures, the magnetoresistance undergoes a dramatic reversal from negative to positive. This reversal points to a potential field-induced phase transition into a collective electronic state driven by enhanced electron-electron interactions. Our findings provide valuable insights into the unconventional chiral transport and emergent correlated quantum states under extreme magnetic fields.

cond-mat.mtrl-sci↗

arXiv2026

CQDC-Warp: GPU-Accelerated Differentiable Quasi-Dynamic Simulation for Contact-Rich Manipulation

Efficient synthesis of contact-rich manipulation trajectories is essential for downstream robot learning. Gradient-based planning offers a promising approach, but jointly optimizing motion and contact decisions remains challenging. Meanwhile, many existing planners run on CPUs with limited throughput. We present CQDC-Warp, a GPU-accelerated differentiable quasi-dynamic simulator for batched model predictive control (MPC). We use it to study contact-guided trajectory optimization, in which intended robot--object contacts are specified before planning and encouraged through soft costs, while the simulated dynamics determine contact geometry and frictional response without a prescribed contact-mode schedule. Across $17{,}408$ trials on the $17$-object Allegro in-hand rotation benchmark, our framework achieves $79.8\%$ success, compared with $73.7\%$ for CFree and $65.2\%$ for IMPACT, and reduces mean trajectory-generation time by factors of $27.0$ and $50.2$, respectively. On a single NVIDIA RTX 4090 with $2{,}048$ parallel worlds, forward and backward throughput reach $2.06\times10^5$ and $2.25\times10^5$ world-steps per second, respectively. Further experiments cover manipulation with LEAP and Shadow hands, environmental contact exploitation, bimanual manipulation, whole-body control, and real-world trajectory replay on a Sharpa hand.

cs.RO↗

arXiv2026

Uni-PNP: Unified Prehensile and Non-prehensile Policy Learning for Generalizable Object Pose Reaching

Object manipulation often requires combining non-prehensile actions with grasping. However, existing approaches typically develop these capabilities separately and integrate them by sequencing distinct primitives through prescribed intermediate objectives, limiting the coordination of contact motions and grasping. Their integration also remains sensitive to object and gripper geometry and depends on particular environmental supports, complicating real-world deployment. We present Uni-PNP, a unified policy that jointly learns prehensile and non-prehensile manipulation for object pose reaching across diverse objects and grippers. To achieve this, we first pretrain a geometry encoder to predict contact-induced object motion using a contact-dynamics dataset constructed from simulated interactions between diverse objects and grippers. We then train the policy in a two-stage reinforcement learning curriculum from graspable objects to ungraspable peers, encouraging the discovery of non-prehensile interactions and their integration with grasping. Experiments show 76.39% success on hundreds of seen objects with grippers, and 57.91% on unseen peers, outperforming the strongest pose-reaching baseline by 20.52 and 17.79 percentage points, respectively. We also analyze emergent behaviors, including nonprehensile and prehensile repositioning, grasp acquisition that requires grasping after tipping, and table-assisted reorientation while grasping, and transfer to the real-world scenario. Project page available at lgswdn.github.io/Uni-PNP.

cs.RO↗

arXiv2026

When and Why Adversarial Training Improves PINNs: A Neural Tangent Kernel Perspective

Physics-informed neural networks (PINNs) are powerful surrogates for differential equations but are notoriously difficult to train due to spectral bias, stiffness, and poor accuracy on high-frequency or multiscale solutions. Adversarial training based on generative adversarial networks (GANs) has recently gained surprisingly strong empirical results in improving training, but the underlying mechanisms remain elusive. To this end, we propose a new analysis framework for adversarially trained PINNs, based on the key observation of how the discriminator in GANs can influence the training dynamics of PINNs. The framework first provides a much needed theoretical grounding to why and when adversarial training is effective in PINNs, then presents a unified analysis of GANs variants in such training, and finally leads to a new, practical, efficient training algorithm for PINNs. Empirical results demonstrate that our method can significantly reduce the pathology of PINNs training, thereby providing better models with superior performances, often several magnitudes more accurate than alternative methods.

cs.LG↗