arXiv subjects

Ming Zhou

Publications and source records attributed to Ming Zhou.

At least 37 records · Page 2Linked to original sources

arXiv2026

LoopCoder-v2: Only Loop Once for Efficient Test-Time Computation Scaling

Looped Transformers scale latent computation by repeatedly applying shared blocks, but sequential looping increases latency and KV-cache memory with the loop count. Parallel loop Transformers (PLT) alleviate this cost through cross-loop position offsets (CLP) and shared-KV gated sliding-window attention, making loop count a practical design choice. We therefore study PLT loop-count selection through a gain--cost view: an extra loop may refine representations, but CLP also introduces a positional mismatch at each loop boundary. We instantiate this study by training LoopCoder-v2, a family of 7B PLT coders with different loop counts, from scratch on 18T tokens, followed by matched instruction tuning and evaluation. Empirically, the two-loop variant delivers broad gains over the non-looped baseline across code generation, code reasoning, agentic software engineering, and tool-use benchmarks, improving SWE-bench Verified from 43.0 to 64.4 points and Multi-SWE from 14.0 to 31.0 points. In contrast, variants with three or more loops regress, revealing a strongly non-monotonic loop-count effect. Our diagnostics show that loop 2 provides the main productive refinement, while later loops yield diminishing, oscillatory updates and reduced representational diversity. Because the CLP-induced mismatch remains roughly fixed as refinement gains shrink, the offset cost increasingly dominates. This gain--cost trade-off explains PLT's saturation at two loops and provides diagnostics for loop-count selection.

cs.LG↗

arXiv2026

Scalable and General Whole-Body Control for Cross-Humanoid Locomotion

Learning-based whole-body controllers have become a key driver for humanoid robots, yet most existing approaches require robot-specific training. In this paper, we study the problem of cross-embodiment humanoid control and show that a single policy can robustly generalize across a wide range of humanoid robot designs with one-time training. We introduce XHugWBC, a novel cross-embodiment training framework that enables generalist humanoid control through: (1) physics-consistent morphological randomization, (2) semantically aligned observation and action spaces across diverse humanoid robots, and (3) effective policy architectures modeling morphological and dynamical properties. XHugWBC is not tied to any specific robot. Instead, it internalizes a broad distribution of morphological and dynamical characteristics during training. By learning motion priors from diverse randomized embodiments, the policy acquires a strong structural bias that supports zero-shot transfer to previously unseen robots. Experiments on twelve simulated humanoids and seven real-world robots demonstrate the strong generalization and robustness of the resulting universal controller.

cs.RO↗

arXiv2026

DanceHMR: Hand-Aware Whole-Body Human Mesh Recovery from Monocular Videos

Monocular video human mesh recovery is essential for digital humans, avatar animation, and embodied simulation, where both temporal stability and expressive whole-body motion are required. Existing video HMR methods produce coherent body motion but often overlook detailed hand articulation, while image-based whole-body methods recover SMPL-X meshes independently per frame, often leading to jittery and inaccurate hand motion. We present a temporally coherent whole-body HMR framework for challenging in-the-wild monocular videos. Our model unifies body context and part-specific hand observations through residual body-hand fusion, enabling stable body motion and detailed hand recovery within a single temporal architecture. We further introduce close-up-aware augmentation to improve robustness under upper-body framing. Experiments on whole-body and body-only benchmarks demonstrate improved hand reconstruction and competitive body accuracy. Our method also produces temporally stable and 2D-consistent SMPL-X motion in challenging real-world videos.

cs.CV↗

arXiv2026

MIRA: Mid-training Rubric Anchoring for Source-Aware Data Selection

Mid-training has become an important stage in modern LLM development, using large-scale curated mixtures to strengthen capabilities before final post-training. Its data selection problem is distinct: the data are optimized under a pretraining-style objective at near-pretraining scale, but are curated toward downstream capabilities and drawn from heterogeneous sources with different formats and training roles. As a result, effective selection requires both scalability and source-adaptive semantic criteria. Existing model-based methods scale well, but provide only implicit quality signals. Semantic selection methods offer stronger judgments, but usually assume fixed rubrics or standardized data formats. To address this mismatch, we propose MIRA, a source-aware filtering framework based on self-anchored rubric discovery. The key idea is to make rubric construction part of data selection: MIRA first discovers what should be evaluated for each source group, then distills those judgments into scalable student scorers for full-corpus filtering. On code-oriented mid-training with 21 sources and 5 source groups, MIRA outperforms selection baselines across nine code benchmarks and matches the full-corpus run while using only half the tokens.

cs.AI↗

arXiv2026

DyGRO-VLA: Cross-Task Scaling of Vision-Language-Action Models via Dynamic Grouped Residual Optimization

Recent progress in Reinforcement Learning (RL) provides a principled approach to optimizing Vision-Language-Action (VLA) models, facilitating a shift from trajectory imitation to active learning in the task environment. Despite improvements in control precision, most RL optimizers remain task-specific, which reduces VLA models from generalist controllers to policies that overfit to a narrow set of tasks. In this study, we conduct an in-depth analysis of this phenomenon and highlight the importance of cross-task feature representations for improving the generalizability of VLA models. Motivated by this finding, we introduce DyGRO-VLA, a two-stage optimization framework that 1) effectively captures cross-task latent representations based on information-theoretic principles, and 2) dynamically refines policy optimization via a mixture-of-RL-residuals. DyGRO-VLA enables the RL optimizer to exploit task-relevant latent information while strategically mitigating adverse interference on the learned representations throughout the optimization process. We evaluate our approach on LIBERO, RoboTwin2 benchmarks, and further validate it on real world, demonstrating consistent improvements over strong baselines under multi-task training and distribution shift.

cs.RO↗

arXiv2026

Discovery and Characterization of White Dwarf-FGK Main-Sequence Binaries within the Optical Main-Sequence Locus

White dwarf main-sequence (WDMS) binaries provide important laboratories for studying binary evolution and the formation of low-mass white dwarfs. In this work, we identify 654 reliable WDMS candidates with FGK-type companions from an initial set of 772 ultraviolet-excess sources, selected using stellar atmospheric parameters from LAMOST spectroscopy and subsequently refined with \textit{Gaia} DR3 astrometry and photometry together with ultraviolet data from \textit{GALEX}. Candidates were selected based on ultraviolet excess relative to the \textit{Gaia} main-sequence locus and refined using isochrone constraints to exclude systems inconsistent with MS companions. Binary spectral energy distribution fitting yields effective temperatures and radii for both components, as well as distance and extinction estimates. The MS companions are dominated by G-type stars ($\sim52\%$), with comparable fractions of F- and K-type companions, and no A-type primaries. Using white-dwarf evolutionary cooling models, we find that the WD components are predominantly low-mass ($M_{\rm WD}\,\sim\,0.2\text{--}0.4\,M_\odot$), including a substantial population of extremely low-mass ($<0.3\,M_\odot$) WDs likely produced through binary interaction. The WDs are generally hot ($\sim1.5\times10^4\,\mathrm{K}$), consistent with the ultraviolet selection bias favoring luminous, large-radius WDs. Multi-epoch LAMOST radial velocities show larger amplitudes than those of a comparison sample of MS stars, supporting the close-binary nature of these systems. Although subject to strong selection effects, the catalog offers a clean and well-characterized sample of FGK+WD binaries.

astro-ph.SR↗

arXiv2026

Characterizing the Instrumental Profile of LAMOST

The instrumental profile (IP) of a telescope is of great significance for spectroscopic analyses, especially for wavelength calibration and stellar parameter measurements. The Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) employs arc lamps for wavelength calibration. These lamps produce sharp emission lines with known wavelengths, and the observed arc lamp spectra can well characterize the IP. However, IPs are influenced by multiple factors, making them difficult to model accurately with traditional methods. Neural networks, which can automatically capture complex patterns and nonlinear features in data, provide a promising approach for high-precision IP measurement. We therefore construct a multi-layer perceptron (MLP) based on The Payne neural network to derive IPs for LAMOST. After training, the model can retrieve the IP for any fiber, at any wavelength, and at any time. We then apply the derived IP to stellar radial velocity (RV) measurements and analyze the impact of different IP center localization methods on the results. Finally, the dispersion of the measured RVs is reduced by approximately 3 km/s. This improvement will facilitate the search for long-period binary stars via RV variations.

astro-ph.IM↗

arXiv2026

InCoder-32B-Thinking: Industrial Code World Model for Thinking

Industrial software development across chip design, GPU optimization, and embedded systems lacks expert reasoning traces showing how engineers reason about hardware constraints and timing semantics. In this work, we propose InCoder-32B-Thinking, trained on the data from the Error-driven Chain-of-Thought (ECoT) synthesis framework with an industrial code world model (ICWM) to generate reasoning traces. Specifically, ECoT generates reasoning chains by synthesizing the thinking content from multi-turn dialogue with environmental error feedback, explicitly modeling the error-correction process. ICWM is trained on domain-specific execution traces from Verilog simulation, GPU profiling, etc., learns the causal dynamics of how code affects hardware behavior, and enables self-verification by predicting execution outcomes before actual compilation. All synthesized reasoning traces are validated through domain toolchains, creating training data matching the natural reasoning depth distribution of industrial tasks. Evaluation on 14 general (81.3% on LiveCodeBench v5) and 9 industrial benchmarks (84.0% in CAD-Coder and 38.0% on KernelBench) shows InCoder-32B-Thinking achieves top-tier open-source results across all domains.GPU Optimization

cs.AR↗

arXiv2026

Quantum algorithms for the fractional Poisson equation via rational approximation

This paper presents a quantum algorithm for solving the fractional Poisson equation $(-Δ)^s u = f$ with $s \in (0,1)$ on bounded domains. The proposed approach combines rational approximation techniques with quantum linear system solvers to achieve exponential quantum advantage. The rational approximation represents the inverse fractional Laplacian as a weighted sum of standard resolvents, transforming the original nonlocal problem into a collection of shifted integer-order partial differential equations. These equations are consolidated into a single large linear system through a modified right-hand side construction that simplifies the quantum implementation. To enable practical implementation, we develop explicit quantum circuits via the Schrödingerization technique, which converts the non-unitary dynamics of the linear system into a higher-dimensional Schrödinger-type equation, allowing the use of standard Hamiltonian simulation. The circuit construction leverages the decomposition of shift operators to realize the discrete Laplacian and employs controlled operations to implement the select oracle. Under finite difference discretization, we provide detailed algorithmic procedures utilizing block-encoding techniques for the coefficient matrices. A comprehensive complexity analysis demonstrates that the quantum algorithm achieves a dependence on the inverse mesh size $h^{-1}$ that is independent of the spatial dimension $d$, in stark contrast to classical methods which suffer from exponential growth in high dimensions. This establishes an exponential quantum advantage for high-dimensional fractional problems, effectively overcoming the curse of dimensionality that limits classical approaches.

quant-ph↗

arXiv2026

InCoder-32B: Code Foundation Model for Industrial Scenarios

Recent code large language models have achieved remarkable progress on general programming tasks. Nevertheless, their performance degrades significantly in industrial scenarios that require reasoning about hardware semantics, specialized language constructs, and strict resource constraints. To address these challenges, we introduce InCoder-32B (Industrial-Coder-32B), the first 32B-parameter code foundation model unifying code intelligence across chip design, GPU kernel optimization, embedded systems, compiler optimization, and 3D modeling. By adopting an efficient architecture, we train InCoder-32B from scratch with general code pre-training, curated industrial code annealing, mid-training that progressively extends context from 8K to 128K tokens with synthetic industrial reasoning data, and post-training with execution-grounded verification. We conduct extensive evaluation on 14 mainstream general code benchmarks and 9 industrial benchmarks spanning 4 specialized domains. Results show InCoder-32B achieves highly competitive performance on general tasks while establishing strong open-source baselines across industrial domains.

cs.SE↗

arXiv2026

Searching for Magnetic White Dwarfs in LAMOST DR10

Magnetic white dwarfs (MWDs) are key to understanding the origin and evolution of magnetic fields in compact stars. While large spectroscopic surveys such as SDSS have greatly expanded the known sample, the potential of LAMOST has not yet been fully explored. Our aim is to identify and characterize isolated MWDs in the LAMOST DR10 database. We cross-matched LAMOST DR10 spectra with white dwarf candidates from Gaia EDR3 and with recent SDSS-based catalogs of MWDs. Zeeman splitting in Balmer and helium absorption lines was used as the primary diagnostic to identify magnetic fields and to estimate their strengths. Reference objects from SDSS catalogs were used to test the detectability of MWDs in LAMOST low-resolution spectra. We identified 63 isolated MWDs in LAMOST DR10, of which 32 are new discoveries. Surface magnetic field strengths were measured from Zeeman splitting, covering a range from a few MG up to several tens of MG. For previously known SDSS MWDs, our LAMOST-based field measurements show mostly agreement with published values. This work demonstrates the capability of LAMOST low-resolution spectroscopy to identify and characterize isolated MWDs. The newly discovered objects expand the known population and provide valuable targets for future high-resolution spectroscopic and polarimetric follow-up studies. Our results highlight the potential of combining LAMOST with Gaia and other large surveys to build a more complete census of MWDs.

astro-ph.SR↗

arXiv2026

Revisiting Very High Energy Gamma-Ray Absorption in Cosmic Propagation under the Combined Effects of Axion-Like Particles and Lorentz Invariance Violation

Very-high-energy (VHE; $E \gtrsim 100$ GeV) gamma rays are expected to experience strong attenuation during cosmological propagation due to electron-positron pair production on the extragalactic background light (EBL). Recent observations of GRB 221009A (z = 0.151), including photons up to $\sim 18$ detected by LHAASO and a $\sim 300\ \mathrm{TeV}$ event reported by Carpet-3, suggest a higher-than-expected transparency of the Universe at extreme energies. These observations cannot be explained by standard EBL absorption alone; moreover, neither Lorentz invariance violation (LIV) nor photon-axion-like particle (ALP) oscillations, when considered in isolation, appear sufficient to account for the survival of such photons over cosmological distances. In this work, we propose a joint propagation scenario that incorporates photon-ALP mixing in astrophysical magnetic fields together with subluminal quadratic LIV corrections to the $γγ$ pair-production threshold. Applying this framework to the broadband gamma-ray spectrum of GRB 221009A, we show that ALPs with coupling ($g_{aγ} = 1.685 \times 10^{-10}\mathrm{GeV}^{-1}$ ) and mass ($m_a = 9.545 \times 10^{-8}\mathrm{eV}$), combined with a quadratic LIV energy scale ($E_{\rm LIV,2} = 1.30 \times 10^{-7} E_{\rm Pl}$) adopted from the literature, can significantly enhance the photon survival probability in the energy range (10\text{-}300) TeV. The resulting enhancement exceeds that obtained from either ALP mixing or LIV effects alone. These results indicate that a combined ALP-LIV scenario may provide a viable interpretation of the extreme-energy gamma-ray observations of GRB 221009A and highlight the potential of VHE gamma-ray measurements as probes of physics beyond the Standard Model.

astro-ph.HE↗

arXiv2026

LightRetriever: A LLM-based Text Retrieval Architecture with Extremely Faster Query Inference

Large Language Models (LLMs)-based text retrieval retrieves documents relevant to search queries based on vector similarities. Documents are pre-encoded offline, while queries arrive in real-time, necessitating an efficient online query encoder. Although LLMs significantly enhance retrieval capabilities, serving deeply parameterized LLMs slows down query inference throughput and increases demands for online deployment resources. In this paper, we propose LightRetriever, a novel LLM-based retriever with extremely lightweight query encoders. Our method retains a full-sized LLM for document encoding, but reduces the workload of query encoding to no more than an embedding lookup. Compared to serving a full LLM on an A800 GPU, our method achieves over 1000x speedup in query encoding and over 10x increase in end-to-end retrieval throughput. Extensive experiments on large-scale retrieval benchmarks show that LightRetriever generalizes well across diverse tasks, maintaining an average of 95% retrieval performance.

cs.IR↗

arXiv2026

RoboStriker: Hierarchical Decision-Making for Autonomous Humanoid Boxing

Achieving human-level competitive intelligence and physical agility in humanoid robots remains a major challenge, particularly in contact-rich and highly dynamic tasks such as boxing. While Multi-Agent Reinforcement Learning (MARL) offers a principled framework for strategic interaction, its direct application to humanoid control is hindered by high-dimensional contact dynamics and the absence of strong physical motion priors. We propose RoboStriker, a hierarchical three-stage framework that enables fully autonomous humanoid boxing by decoupling high-level strategic reasoning from low-level physical execution. The framework first learns a comprehensive repertoire of boxing skills by training a single-agent motion tracker on human motion capture data. These skills are subsequently distilled into a structured latent manifold, regularized by projecting the Gaussian-parameterized distribution onto a unit hypersphere. This topological constraint effectively confines exploration to the subspace of physically plausible motions. In the final stage, we introduce Latent-Space Neural Fictitious Self-Play (LS-NFSP), where competing agents learn competitive tactics by interacting within the latent action space rather than the raw motor space, significantly stabilizing multi-agent training. Experimental results demonstrate that RoboStriker achieves superior competitive performance in simulation and exhibits sim-to-real transfer. Our website is available at RoboStriker.

cs.RO↗

arXiv2026

Toward genuine efficiency and cluster robustness of preconditioned CG-like eigensolvers

The performance of eigenvalue problem solvers (eigensolvers) depends on various factors such as preconditioning and eigenvalue distribution. Developing stable and rapidly converging vectorwise eigensolvers is a crucial step in improving the overall efficiency of their blockwise implementations. The present paper is concerned with the locally optimal block preconditioned conjugate gradient (LOBPCG) method for Hermitian eigenvalue problems, and motivated by two recently proposed alternatives for its single-vector version LOPCG. A common basis of these eigensolvers is the well-known CG method for linear systems. However, the optimality of CG search directions cannot perfectly be transferred to CG-like eigensolvers. In particular, while computing clustered eigenvalues, LOPCG and its alternatives suffer from frequent delays, leading to a staircase-shaped convergence behavior which cannot be explained by the existing estimates. Keeping this in mind, we construct a class of cluster robust vector iterations where LOPCG is replaced by asymptotically equivalent two-term recurrences and the search directions are timely corrected by selecting a far previous iterate as augmentation. The new approach significantly reduces the number of required steps and the total computational time.

math.NA↗

arXiv2025

H-Zero: Cross-Humanoid Locomotion Pretraining Enables Few-shot Novel Embodiment Transfer

The rapid advancement of humanoid robotics has intensified the need for robust and adaptable controllers to enable stable and efficient locomotion across diverse platforms. However, developing such controllers remains a significant challenge because existing solutions are tailored to specific robot designs, requiring extensive tuning of reward functions, physical parameters, and training hyperparameters for each embodiment. To address this challenge, we introduce H-Zero, a cross-humanoid locomotion pretraining pipeline that learns a generalizable humanoid base policy. We show that pretraining on a limited set of embodiments enables zero-shot and few-shot transfer to novel humanoid robots with minimal fine-tuning. Evaluations show that the pretrained policy maintains up to 81% of the full episode duration on unseen robots in simulation while enabling few-shot transfer to unseen humanoids and upright quadrupeds within 30 minutes of fine-tuning.

cs.RO↗

arXiv2025

Maglev-Pentabot: Magnetic Levitation System for Non-Contact Manipulation using Deep Reinforcement Learning

Non-contact manipulation has emerged as a transformative approach across various industrial fields. However, current flexible 2D and 3D non-contact manipulation techniques are often limited to microscopic scales, typically controlling objects in the milligram range. In this paper, we present a magnetic levitation system, termed Maglev-Pentabot, designed to address this limitation. The Maglev-Pentabot leverages deep reinforcement learning (DRL) to develop complex control strategies for manipulating objects in the gram range. Specifically, we propose an electromagnet arrangement optimized through numerical analysis to maximize controllable space. Additionally, an action remapping method is introduced to address sample sparsity issues caused by the strong nonlinearity in magnetic field intensity, hence allowing the DRL controller to converge. Experimental results demonstrate flexible manipulation capabilities, and notably, our system can generalize to transport tasks it has not been explicitly trained for. Furthermore, our approach can be scaled to manipulate heavier objects using larger electromagnets, offering a reference framework for industrial-scale robotic applications.

cs.RO↗

arXiv2025

A Vision-Language-Action-Critic Model for Robotic Real-World Reinforcement Learning

Robotic real-world reinforcement learning (RL) with vision-language-action (VLA) models is bottlenecked by sparse, handcrafted rewards and inefficient exploration. We introduce VLAC, a general process reward model built upon InternVL and trained on large scale heterogeneous datasets. Given pairwise observations and a language goal, it outputs dense progress delta and done signal, eliminating task-specific reward engineering, and supports one-shot in-context transfer to unseen tasks and environments. VLAC is trained on vision-language datasets to strengthen perception, dialogic and reasoning capabilities, together with robot and human trajectories data that ground action generation and progress estimation, and additionally strengthened to reject irrelevant prompts as well as detect regression or stagnation by constructing large numbers of negative and semantically mismatched samples. With prompt control, a single VLAC model alternately generating reward and action tokens, unifying critic and policy. Deployed inside an asynchronous real-world RL loop, we layer a graded human-in-the-loop protocol (offline demonstration replay, return and explore, human guided explore) that accelerates exploration and stabilizes early learning. Across four distinct real-world manipulation tasks, VLAC lifts success rates from about 30\% to about 90\% within 200 real-world interaction episodes; incorporating human-in-the-loop interventions yields a further 50% improvement in sample efficiency and achieves up to 100% final success.

cs.RO↗