arXiv subjects

Ke He

Publications and source records attributed to Ke He.

At least 19 recordsLinked to original sources

arXiv2026

Gate-tunable superconductor-semiconductor coupling and its impact on zero-energy modes in PbTe-Pb nanowires

Superconductor-semiconductor coupling in hybrid nanowires plays a pivotal role in Majorana devices. Here, we demonstrate this tunable coupling in PbTe-Pb nanowires and its impact on the induced gap and zero-bias peaks (ZBPs). In the strong coupling regime, the device reveals a clean superconducting gap, where robust and non-robust ZBPs are observed in distinct magnetic field ranges. Conversely, increasing the gate voltage drives the system toward the weak coupling regime, where the gap becomes narrow and soft. Consequently, robust ZBPs disappear while non-robust ones persist. Our results establish a direct link between tunable coupling and robust ZBPs, providing guidelines on engineering Majorana zero modes and topological qubits.

cond-mat.mes-hall↗

arXiv2026

Generative Actor-Critic with Soft Bridge Policies

Expressive generative policies such as diffusion and flow models are appealing for maximum-entropy (MaxEnt) online reinforcement learning because of their ability to model multimodal and highly non-Gaussian action distributions. However, training effective soft generative policies faces two obstacles that often arise together. First, marginal action densities are often unavailable, so existing methods typically rely on entropy bounds, heuristic proxies or approximations. Second, iterative shared-parameter samplers raise inference cost and require backpropagation through time over repeated network evaluations, increasing memory cost and potentially destabilizing policy optimization. These obstacles motivate us to seek a generative actor-critic method that exposes a tractable soft objective while generating each action in just one network evaluation. To this end, we propose soft generative actor-critic (SoftGAC), whose actor is a single stochastic feedforward network with internal Gaussian bridge layers. This explicit bridge structure allows us to lift the soft objective to path space, yielding an analytical path-KL regularizer against a high-entropy reference process for both actor improvement and the soft Bellman bootstrap, without estimating the endpoint entropy. Crucially, the actor generates each action in only one network evaluation under a total parameter budget comparable to strong baselines, substantially reducing inference cost relative to the existing iterative diffusion and flow policies. It also avoids long unrolled backpropagation through time. Across challenging continuous-control benchmarks, our method achieves competitive or superior returns compared with strong diffusion and flow policies, demonstrating a clear advantage in the compute-return tradeoff.

cs.LG↗

arXiv2026

Radioactive Feedback: Tracing Models Trained by Label-Only LLM Supervision

Conventional distillation transfers capability through teacher-generated demonstrations. LLM-as-a-Judge knowledge distillation offers a distinct and increasingly practical alternative. In the label-only variant studied here, the student generates every response and improves from scalar scores alone. Prior work shows that this non-demonstrative feedback can improve student capability, but it creates a fundamentally different provenance problem. No teacher-generated answer enters the student's training set, so any evaluator-specific trace must pass through a low-bandwidth reward signal rather than copied text. Existing watermarks and distillation detectors do not cover this score-only channel. We introduce Radioactive Feedback, which encodes a registered pattern of behavioral preferences in the rewards themselves. The evaluator applies bounded pointwise adjustments to quality-qualified responses, allowing repeated optimization to transfer the pattern to a student trained by another party. Experiments on code and mathematics show that the registered preferences transfer through scores alone and across datasets, enabling black-box detection against a fixed reference pool.

cs.AI↗

arXiv2026

Semantic Modality Compensation for Unsupervised Visible-Infrared Person Re-identification under Unpaired Settings

Unsupervised visible-infrared person re-identification (USL-VI-ReID) learns person representations that can be compared across modalities without identity annotations. In the unpaired setting, however, identity correspondences between modalities are often incomplete, leaving many identities without an observed counterpart in the other modality. Existing unpaired methods bridge this gap by generating or mapping features for the other modality, mainly by exploiting the statistics of visual features without explicitly separating content that is discriminative for identity from style that is specific to modality. Consequently, the generated features may distort identity cues or inherit bias from the source modality, undermining the reliability of supervision across modalities. We formulate unpaired learning across modalities as a semantic compensation problem and propose Semantic Modality Compensation (SMC), a framework based on prompt composition that decouples identity semantics from modality style within a shared visual semantic space. SMC first constructs a discriminative ReID space through augmented dual contrastive learning, yielding pseudo labels, cluster prototypes, and memory banks for each modality. It then learns visible and infrared modality prompts in the CLIP semantic space and maps clusters obtained from pseudo labels to identity semantic tokens. For each cluster lacking a reliable match in the other modality, SMC combines its identity token with the prompt for the target modality to synthesize a semantic counterpart in the missing modality. The synthesized counterpart is then projected back into the ReID space and injected into a compensation memory through confidence gating. Extensive experiments under both paired and unpaired settings demonstrate that SMC consistently outperforms state-of-the-art methods, with particularly large gains when identity mismatch is severe.

cs.CV↗

arXiv2026

OneWorld: Learning Consistent Physics Across Actions in World Models

Action-conditioned video world models aim to predict scene evolution under different actions, a capability that is essential for reliable planning, decision-making, and interaction in dynamic environments. However, futures generated independently from the same initial scene may each appear plausible while implying incompatible physical properties, such as friction or mass. This inconsistency can lead to contradictory predictions across interventions, making it difficult for the model to maintain a coherent understanding of the underlying world and limiting its reliability for planning and decision-making. To address these issues, we propose OneWorld, a shared-mechanism counterfactual generation framework that jointly models multiple action-conditioned futures under a common latent physical mechanism. A physical mechanism interpreter first infers a distribution over latent mechanisms from each action-outcome branch. These distributions are then aggregated into shared-world evidence, which captures whether the branches admit a common physical explanation while accounting for uncertainty in less informative branches. This evidence constrains flow training and guides sampling, encouraging consistency in the underlying physical mechanism while preserving the distinct outcomes induced by different actions. We further introduce a multi-intervention evaluation protocol in controlled environments, following the interaction settings of ACWM-Phys, to assess whether generated futures can be jointly explained by the same physical parameters, alongside standard measures of single-rollout prediction quality. Experiments in these environments show that OneWorld improves cross-intervention physical consistency while maintaining competitive single-rollout prediction quality.

cs.CV↗

arXiv2026

Presence versus absence of charging energies in PbTe quantum dots

Charging energy ($E_C$) is essential in quantum dot (QD) devices. Previous studies on PbTe QDs have reported both the presence and absence of $E_C$. To resolve this ambiguity, we vary the QD size, i.e. the cross-sectional area of PbTe nanowires, and track the evolution of $E_C$. For large crosssectional areas ($\sim$ 16000 nm$^2$), the PbTe QDs exhibit no measurable $E_C$, while quantized levels are well resolved. Decreasing this area successively to 5000, 1500, and 460 nm$^2$, $E_C$ becomes finite and increases to 80, 160, and 210 $μ$eV, respectively. We further demonstrate the strong tunability of local gates, which can tune the PbTe device from the QD regime to the regime of ballistic transport. These results address concerns regarding the large dielectric constant of PbTe and provide key insights in engineering advanced PbTe quantum devices.

cond-mat.mes-hall↗

arXiv2026

Probing PbTe-Pb nanowire devices with radio-frequency reflectometry

We report the implementation of radio-frequency (rf) reflectometry on selective-area-grown PbTe-Pb nanowire devices on a CdTe substrate. These nanowires are predicted to host Majorana zero modes. We demonstrate the compatibility of the rf technique, including both resistive and capacitive sensing, with these nanowires. The effect of dielectric loss from the CdTe substrate is quantitatively characterized. Furthermore, the feasibility of rf reflectometry is verified under finite magnetic fields where zero-energy modes can emerge. Our results establish the fast control of PbTe quantum devices, paving the way for their applications in topological quantum computation.

cond-mat.mes-hall↗

arXiv2026

Reversible tuning of magnetic order and intrinsic superconductivity in strained FeTe films via stoichiometry control

FeTe is a prototypical parent compound of iron-based superconductors. While bulk FeTe is non-superconducting with a long-range bicollinear antiferromagnetic order, superconductivity has been achieved in thin films. However, the approaches usually involve complex oxygen incorporation or interfacial effects, the microscopic mechanisms of which remain elusive. Here, we prepare high-purity, bare FeTe thin films on SrTiO3 and investigate their magnetic and superconducting states combining both microscopic and macroscopic characterizations. By reducing the interstitial Fe impurities, we successfully suppress the long-range antiferromagnetic order, enhance the quasiparticle coherence and induce superconductivity at ~10 K. Moreover, this process is readily reversible by tuning the Fe concentration. Our findings reveal that precise stoichiometric control is sufficient to induce intrinsic superconductivity in strained FeTe thin films. This work provides insights into the competition between magnetism and superconductivity in iron chalcogenides, and supplies methods for developing stable, high-purity superconducting FeTe films.

cond-mat.supr-con↗

arXiv2026

See Before You Code: Learning Visual Priors for Spatially Aware Educational Animation Generation

Large language models can generate executable code for educational animations, but the resulting renders often exhibit visual defects, including element overlap, misalignment, and broken animation continuity. These defects cannot be reliably detected from the code alone and become apparent only after execution. We formalize this problem as render-feedback-aware constrained code generation: given a natural language specification, the model must generate executable code whose rendered output satisfies structured quality criteria that can be evaluated only after rendering. To address this problem, we introduce OmniManim, a render-feedback-aware educational animation generation framework built around a shared scene state, explicit visual planning, structured post-render diagnostics, and localized repair. Within OmniManim, the Vision Agent is a task-specific visual planning module: it predicts sparse keyframe layouts with coarse-to-fine bounding-box denoising and optimizes an interpolation-aware objective to reduce intermediate-frame failures induced by downstream animation interpolation. We further construct two datasets, ManimLayout-1K and EduRequire-500, and provide a reproducible evaluation protocol covering executability, instructional quality, visual quality, and efficiency. On EduRequire-500, OmniManim improves measured render quality over both single-model baselines and existing multi-agent frameworks. Systematic ablation studies further verify that explicit visual planning, especially its coarse spatial prior, bounding-box refinement, and interpolation-aware optimization, is central to these gains.

cs.AI↗

arXiv2026

Strong enhancement of g-factor in PbTe-Pb hybrid nanowires

We report large Lande g-factors observed in PbTe-Pb hybrid nanowires. The g-factor can reach 83, significantly larger than those in bare PbTe nanowires (typically below 20). We attribute this enhancement to orbital effects in the superconducting film, particularly when the magnetic field is nearly perpendicular to the Pb film. This enhancement is beneficial for the search for topological superconductivity by reducing the critical magnetic field required for the phase transition.

cond-mat.mes-hall↗

arXiv2026

Excess photon-assisted noise of Majorana and Andreev bound states

Photon-assisted tunneling arises under an ac bias, with the drive frequency setting the photon energy. The excess photon-assisted noise is defined as the difference between the shot noise under a combined dc and ac bias and that under a dc bias alone. We investigate this quantity in tunneling into Majorana or Andreev bound states, which are of great interest in the search for topological superconductors. Under a harmonic bias $V(t)=V_\mathrm{dc}[1-\cos(Ωt)]$, the excess photon-assisted noise exhibits distinct behaviors: for Majorana or quasi-Majorana bound states, it undergoes multiple sign reversals as $V_\mathrm{dc}$ increases and vanishes at nonzero integer values of $eV_\mathrm{dc}/Ω$ (with $e$ the elementary charge), whereas for zero-energy Andreev bound states--particularly those producing nearly quantized zero-bias conductance peaks--it remains strictly negative over the entire $V_\mathrm{dc}$ range.

cond-mat.mes-hall↗

arXiv2025

SCA-LLM: Spectral-Attentive LLM-Based Wireless World Modeling for Agentic Communications

Future AI-native wireless networks are moving from reactive optimization to agentic decision-making that can sense, predict, and plan under fast-varying channels. This calls for wireless world models that can predict and roll out channel dynamics, for which multi-step channel state information (CSI) prediction offers a practical short-horizon look-ahead. Recent advances in foundation sequence models further motivate large language models (LLMs) as general-purpose dynamics learners when suitably adapted to non-text time-series signals. However, bridging CSI to LLMs is non-trivial because an effective adapter must expose informative spectral and temporal evolution patterns, while prior designs provide limited inductive bias to capture such channel structures. To this end, we propose SCA-LLM, a spectral-attentive LLM-based wireless world modeling framework that bridges CSI to LLMs via a spectral-channel attention (SCA) adapter. Specifically, the SCA adapter performs multi-spectral representation learning to extract informative channel features and align CSI with the LLM's sequence modeling capability, enabling parameter-efficient adaptation while keeping the LLM backbone largely frozen. Extensive simulations show that SCA-LLM achieves state-of-the-art prediction performance and strong zero-shot generalization, yielding up to -2.4 dB normalized mean squared error (NMSE) advantage over the previous LLM based method. Our ablation studies further confirm the effectiveness of the proposed SCA adapter in mitigating domain mismatch.

cs.IT↗

arXiv2025

QoS-Aware Hierarchical Reinforcement Learning for Joint Link Selection and Trajectory Optimization in SAGIN-Supported UAV Mobility Management

Due to the significant variations in unmanned aerial vehicle (UAV) altitude and horizontal mobility, it becomes difficult for any single network to ensure continuous and reliable threedimensional coverage. Towards that end, the space-air-ground integrated network (SAGIN) has emerged as an essential architecture for enabling ubiquitous UAV connectivity. To address the pronounced disparities in coverage and signal characteristics across heterogeneous networks, this paper formulates UAV mobility management in SAGIN as a constrained multi-objective joint optimization problem. The formulation couples discrete link selection with continuous trajectory optimization. Building on this, we propose a two-level multi-agent hierarchical deep reinforcement learning (HDRL) framework that decomposes the problem into two alternately solvable subproblems. To map complex link selection decisions into a compact discrete action space, we conceive a double deep Q-network (DDQN) algorithm in the top-level, which achieves stable and high-quality policy learning through double Q-value estimation. To handle the continuous trajectory action space while satisfying quality of service (QoS) constraints, we integrate the maximum-entropy mechanism of the soft actor-critic (SAC) and employ a Lagrangian-based constrained SAC (CSAC) algorithm in the lower-level that dynamically adjusts the Lagrange multipliers to balance constraint satisfaction and policy optimization. Moreover, the proposed algorithm can be extended to multi-UAV scenarios under the centralized training and decentralized execution (CTDE) paradigm, which enables more generalizable policies. Simulation results demonstrate that the proposed scheme substantially outperforms existing benchmarks in throughput, link switching frequency and QoS satisfaction.

eess.SP↗

arXiv2025

Magnetic properties of molecular beam epitaxy-grown ultrathin Cr2Ge2Te6 films down to monolayer limit on Si substrates

Cr2Ge2Te6, a prototypical van der Waals ferromagnetic semiconductor, have attracted significant interest for its potential applications in high-performance spintronics. However, the magnetic ground state of monolayer Cr2Ge2Te6 remains elusive due to fragile and irregular-shaped thin flake samples with weak magnetic signals. Here, we successfully grow uniform ferromagnetic Cr2Ge2Te6 films down to monolayer by molecular beam epitaxy. By exploiting a self-limiting growth mode, we achieve synthesis of uniform monolayer Cr2Ge2Te6 films across entire millimeter-scale Si substrates. Through a combination of superconducting quantum interference device magnetometry and anomalous Hall effect measurements, we establish that monolayer Cr2Ge2Te6 exhibits intrinsic ferromagnetism with perpendicular magnetic anisotropy below ~10 K, albeit with strong magnetic fluctuations characteristic of its two-dimensional nature. Furthermore, a systematic thickness-dependent study reveals a crossover from this fluctuation-dominated two-dimensional magnetism turns into conventional long-range ferromagnetic order as the film thickness increases. Our work not only definitively establishes the intrinsic ferromagnetic ground state of monolayer Cr2Ge2Te6, but also provides a scalable, silicon-compatible route for preparing the two-dimensional magnet for future spintronic or quantum devices.

cond-mat.mtrl-sci↗

arXiv2025

Introducing LongCat-Flash-Thinking: A Technical Report

We present LongCat-Flash-Thinking, an efficient 560-billion-parameter open-source Mixture-of-Experts (MoE) reasoning model. Its advanced capabilities are cultivated through a meticulously crafted training process, beginning with long Chain-of-Thought (CoT) data cold-start and culminating in large-scale Reinforcement Learning (RL). We first employ a well-designed cold-start training strategy, which significantly enhances the reasoning potential and equips the model with specialized skills in both formal and agentic reasoning. Then, a core innovation is our domain-parallel training scheme, which decouples optimization across distinct domains (e.g., STEM, Code, Agentic) and subsequently fuses the resulting expert models into a single, nearly Pareto-optimal model. This entire process is powered by our Dynamic ORchestration for Asynchronous rollout (DORA) system, a large-scale RL framework that delivers a greater than threefold training speedup over synchronous methods on tens of thousands of accelerators. As a result, LongCat-Flash-Thinking achieves state-of-the-art performance among open-source models on a suite of complex reasoning tasks. The model exhibits exceptional efficiency in agentic reasoning, reducing average token consumption by 64.5% (from 19, 653 to 6, 965) on AIME-25, without degrading task accuracy. We release LongCat-Flash-Thinking to promote further advances in reasoning systems and agentic AI research.

cs.AI↗

arXiv2025

Asynchronous Risk-Aware Multi-Agent Packet Routing for Ultra-Dense LEO Satellite Networks

The rise of ultra-dense LEO constellations creates a complex and asynchronous network environment, driven by their massive scale, dynamic topologies, and significant delays. This unique complexity demands an adaptive packet routing algorithm that is asynchronous, risk-aware, and capable of balancing diverse and often conflicting QoS objectives in a decentralized manner. However, existing methods fail to address this need, as they typically rely on impractical synchronous decision-making and/or risk-oblivious approaches. To tackle this gap, we introduce PRIMAL, an event-driven multi-agent routing framework designed specifically to allow each satellite to act independently on its own event-driven timeline, while managing the risk of worst-case performance degradation via a principled primal-dual approach. This is achieved by enabling agents to learn the full cost distribution of the targeted QoS objectives and constrain tail-end risks. Extensive simulations on a LEO constellation with 1584 satellites validate its superiority in effectively optimizing latency and balancing load. Compared to a recent risk-oblivious baseline, it reduces queuing delay by over 70%, and achieves a nearly 12 ms end-to-end delay reduction in loaded scenarios. This is accomplished by resolving the core conflict between naive shortest-path finding and congestion avoidance, highlighting such autonomous risk-awareness as a key to robust routing.

cs.NI↗

arXiv2025

LongCat-Flash Technical Report

We introduce LongCat-Flash, a 560-billion-parameter Mixture-of-Experts (MoE) language model designed for both computational efficiency and advanced agentic capabilities. Stemming from the need for scalable efficiency, LongCat-Flash adopts two novel designs: (a) Zero-computation Experts, which enables dynamic computational budget allocation and activates 18.6B-31.3B (27B on average) per token depending on contextual demands, optimizing resource usage. (b) Shortcut-connected MoE, which enlarges the computation-communication overlap window, demonstrating notable gains in inference efficiency and throughput compared to models of a comparable scale. We develop a comprehensive scaling framework for large models that combines hyperparameter transfer, model-growth initialization, a multi-pronged stability suite, and deterministic computation to achieve stable and reproducible training. Notably, leveraging the synergy among scalable architectural design and infrastructure efforts, we complete model training on more than 20 trillion tokens within 30 days, while achieving over 100 tokens per second (TPS) for inference at a cost of \$0.70 per million output tokens. To cultivate LongCat-Flash towards agentic intelligence, we conduct a large-scale pre-training on optimized mixtures, followed by targeted mid- and post-training on reasoning, code, and instructions, with further augmentation from synthetic data and tool use tasks. Comprehensive evaluations demonstrate that, as a non-thinking foundation model, LongCat-Flash delivers highly competitive performance among other leading models, with exceptional strengths in agentic tasks. The model checkpoint of LongCat-Flash is open-sourced to foster community research. LongCat Chat: https://longcat.ai Hugging Face: https://huggingface.co/meituan-longcat GitHub: https://github.com/meituan-longcat

cs.CL↗

arXiv2025

Laughlin charge pumping from interplay of chiral Dirac and chiral Majorana modes

Laughlin charge pumping has provided critical insights into the topological classification of individual materials, but remains largely unexplored in topological junctions. We explore Laughlin charge pumping in junctions composed of a chiral topological superconductor sandwiched between two quantum anomalous Hall insulators, driven by an adiabatically varying magnetic flux. Here, charge pumping can be mediated merely by chiral Dirac modes or by the interplay of chiral Dirac and chiral Majorana modes (CMMs). In the former case, a variation of one magnetic flux quantum induces the pumping of a unit charge, as the chiral Dirac mode accumulates the full flux-induced phase. In contrast, in the latter case, pumping a unit charge requires a variation of fractional magnetic flux quanta, determined by the device geometry and the parity of the number of enclosed superconducting vortices. This unique feature results from the charge-neutral and zero-momentum nature of zero-energy CMMs. Our work offers an experimentally viable pathway toward detecting CMMs and could also inspire further research into Laughlin charge or spin pumping in diverse topological junctions, which are now within experimental reach.

cond-mat.mes-hall↗