arXiv subjects

Lei Zhang

Publications and source records attributed to Lei Zhang.

At least 55 records · Page 3Linked to original sources

arXiv2026

Flash EQ-Linear: Accelerating Equivariant Linear Layers via Group-wise Discrete Fourier Transform

Equivariant networks embed geometric symmetries as structural priors through weight sharing, achieving remarkable parameter efficiency across vision tasks. However, this parameter efficiency does not translate into compute efficiency: most existing implementations unroll the structured weights into dense matrices and dispatch them to generic dense kernels, so an equivariant layer costs no fewer MACs than its non-equivariant counterpart. In this paper, we observe that the equivariant linear (EQ-Linear) layer---the most fundamental and frequently used module in modern equivariant architectures---is essentially a circular convolution along the group dimension composed with a linear transform along the channel dimension. Building on this observation, we propose Flash EQ-Linear, an exact acceleration algorithm that reduces the cost to $2(T-1)/T^2$ of the original dense formulation ($T$ is the equivariant group size) by combining the Fourier convolution theorem along the group dimension with the conjugate symmetry of the real DFT. To translate these computational savings into wall-clock speedups, we further develop dedicated CUDA kernels for the $\mathrm{p}4$ group. At the operator level, Flash EQ-Linear achieves up to $2.1\times$ forward speedup over PyTorch's highly optimized F.linear; at the network level, Flash EQ-ViT achieves up to ${1.7\times}$ end-to-end speedup over both equivariant and non-equivariant baselines. As an operator-level acceleration algorithm, Flash EQ-Linear provides plug-and-play acceleration for diverse pretrained equivariant models, including EQ-ViT, EQ-Swin, EQ-VMamba, and EQ-INR, without retraining or architectural changes. Code is available at https://github.com/zhongchenzhao/FlashEQLinear.

cs.CV↗

arXiv2026

PIVOT: Pivot-Aware On Policy Self Distillation for Multi-Turn VLM Agents

Reinforcement learning with verifiable rewards (RLVR) via Group-Relative Policy Optimization (GRPO) is widely used for multi-turn VLM agent training, yet it suffers from zero-gradient silence on uniform failures and coarse episode-level credit assignment. While On-Policy Distillation (OPD) and On-Policy Self-Distillation (OPSD) mitigate sparse rewards using hindsight information, their underlying mechanisms remain poorly understood. Through controlled counterfactual rollback probes across five multi-turn VLM agent benchmarks, we reveal that performance gains in OPSD/OPD are largely driven by physical state rollback at the pivot step, defined as the first unrecoverable action without remaining step budget. However, physical state rollbacks are computationally prohibitive and infeasible in real-world environments. To bridge this gap, we present Pivot-Aware Internalized Visual On-Policy Training (PIVOT), an RL framework that internalizes pivot localization and state restoration directly into token-level parameter updates, eliminating environment rollbacks during RL training and additional skill hints at test time. PIVOT unifies three functional roles within a single architecture: a failure Analyzer non-invasively localizes the pivot step and diagnoses failure modes from visual trajectory collages and action logs; a detached Teacher re-scores failed tokens under this privileged diagnostic context; and a Student optimizes joint GRPO and confidence-gated OPD objectives. At test time, both Teacher and Analyzer branches are stripped. Evaluated on five multi-turn VLM agent tasks across cognitive grid puzzles, 3D embodied control and navigation, and generative reasoning, PIVOT achieves 0.90 overall accuracy on Qwen2.5-VL-3B (+8% over SFT+GRPO baseline and +5% over previous SOTA) and scales to 0.92 on Qwen3-VL-2B (+12% over SFT+GRPO baseline).

cs.CV↗

arXiv2026

LAGO: Language-Guided Adaptive Object-Region Focus for Zero-Shot Visual-Text Alignment

Zero-shot visual recognition often depends on localized parts, attributes, and textures that are diluted in global image representations. Existing training-free localized methods alleviate this limitation by evaluating multiple image regions, but their candidate locations are typically generated before class semantics determine where to look, restricting language to post-hoc scoring of already sampled evidence. We introduce LAGO (LAnguage-Guided adaptive Object-region focus), which reframes localized recognition as language-guided directed region discovery. LAGO first establishes a compact, class-agnostic object-centric visual initialization, then constructs a soft semantic prototype from the intermediate prediction to actively steer subsequent search toward class-relevant evidence during inference. This object-first, language-second design addresses the circular dependence between recognizing a class and locating its supporting evidence, while preserving complementary local, contextual, and global cues. Across standard and distribution-shift benchmarks, LAGO transfers with a single fixed configuration and uses candidate regions more effectively under matched budgets. Controlled experiments further isolate the proposed mechanism, showing that allowing language to alter the search trajectory retrieves more relevant regions and improves recognition over applying the same semantic signal only after candidate generation.

cs.CV↗

arXiv2026

Separating Diagnosis from Disease Representation: Dual-View EEG Learning with Neural-Dynamics-Guided Deformation

Electroencephalography (EEG)-based closed-loop neuromodulation calls for a subject-specific structured state, as opposed to a single disease probability, specifying which brain regions are deviant, at which frequencies, and at which lags. Sensor-space models keep the strongest diagnostic evidence without anatomy, source-space models give anatomy at a loss of predictive signal, and post-hoc attributions stay outside the prediction. We separate the two instead of forcing them into one representation, and propose DMD-EEG (Dual-view Multiscale Deformation for EEG), which keeps a fixed scalp spectral expert for diagnosis and models the source-space disease-related representation as a low-rank, sparse, iterative deformation of a healthy neural-dynamics prior in a $46$-region-of-interest (ROI) $\times$ $5$-frequency $\times$ $4$-lag (autocorrelation-timescale) space. The two experts meet only at a fixed decision level, so the source state is architecturally separate from the scalp expert. Across major depressive disorder (MDD), first-episode psychosis (FEP), and Parkinson's disease (PD), decision-level fusion matches the strongest single expert on MDD and FEP and exceeds the source branch on PD. On FEP the source expert is the strongest branch, the task where the deformation contributes most. The source state is an explicit ROI-frequency-lag attribution defined in a shared source coordinate system across montages, which we treat as an anatomically-coordinated predictive representation whose coordinates are directly readable and hypothesis-generating. The highest-saliency coordinates align with established disease circuitry (fronto-limbic-temporal regions in MDD, motor-cortex beta in PD), and the MDD state transfers by rank to an unseen cohort recorded with a different montage.

cs.AI↗

arXiv2026

Synchronized Solutions for Liouville Systems with Sign-Changing Coefficients

We consider singular Liouville systems on closed Riemann surfaces with a common sign-changing coefficient. Direct continuation from a positive coefficient is obstructed by the unavoidable appearance of degenerate zeroes along such a homotopy. We instead identify an intrinsic ray in the parameter space on which the system admits synchronized solutions and reduces exactly to a scalar sign-changing mean-field equation. Combining this reduction with the variational existence theory of De Marchis--López-Soriano--Ruiz and the nodal singularity compactness theorem of Wu, we obtain existence for every noncritical parameter on this ray. Positive singular sources are allowed on the nodal curve. In particular, when the surface has nonpositive Euler characteristic and every negative component of the coefficient is a disk, the topological hypothesis required by the scalar min--max construction is automatic.

math.AP↗

arXiv2026

Selective Amortization of Full-Budget Counterfactual Reasoning for Visual Token Communication

Generative image communication transmits compact semantic tokens under a limited packet budget, where token selection directly affects the final reconstruction quality after the complete packet is decoded. However, accurately estimating the terminal value of every candidate token requires repeated receiver-side reconstruction, resulting in substantial encoder-side computation. To address this problem, we propose ACV-Gate, an adaptive candidate evaluation framework that learns to approximate full-budget counterfactual evaluation and selectively assigns exact evaluations to the most informative candidates. Specifically, a set-aware student is trained using terminal advantages and regrets to predict candidate rankings directly, while a selective refinement mechanism evaluates only a bounded candidate set containing both Local-MDL and direct actions; cost-based thresholds further enable explicit control of the average evaluation workload. Experiments on CIFAR-10 show that ACV-Gate consistently improves reconstruction quality while substantially reducing candidate evaluations; at 0.20 bpp, the primary adaptive configuration improves PSNR over LocalMDL by 0.636 dB with only 2.13 candidate evaluations per image, corresponding to 27.60% of the calls required by the Exact-Full expert. Matched-candidate comparisons, synchronized GPU measurements, and evaluations on STL-10 and 384 *384 scale transfer further demonstrate consistent quality computation trade-offs, with particularly pronounced gains at low bit rates. These results show that combining terminal-value learning with selective candidate evaluation provides an effective and controllable mechanism for allocating encoder computation in packet-constrained generative image communication.

cs.AI↗

arXiv2026

Benchmarking the Limits of In-Context Reinforcement Learning for Ad-Hoc Teamwork

In-Context Reinforcement Learning (ICRL) has enabled foundation agents to adapt instantaneously to novel tasks, yet its efficacy in Ad-Hoc Teamwork (AHT)-where coordination with unknown partners is required-remains unexplored. To rigorously evaluate this, we introduce a large-scale benchmark ICRL4AHT, built upon a high-throughput JAX implementation of Overcooked-V2. Our benchmark includes a large, diverse teammate suite spanning both RL and heuristic policies, enabling controlled train-test shifts, and provides a reproducible end-to-end pipeline for teammate generation, learning-history collection, dataset construction, and online multi-episode evaluation. We evaluate representative history-conditioned ICRL algorithms, including Algorithm Distillation (AD) and Decision-Pretrained Transformer (DPT), across millions of transitions. Results reveal notable limitations: contrary to their success in single-agent domains, these baselines fail to exhibit robust test-time adaptation in multi-agent settings. Specifically, these methods frequently underperform random baselines across both unseen teammate and unseen layout tracks, with no clear in-context improvement over long horizons. These findings highlight the challenges of strategic inference under partial observability within the OvercookedV2 AHT protocol, establishing our benchmark as a critical testbed for next-generation coordination algorithms.

cs.AI↗

arXiv2026

Dynamic Generalized Gromov-Wasserstein Optimal Transport

Gromov--Wasserstein optimal transport (GW-OT) extends classical optimal transport by introducing structure-aware transport cost. This is particularly relevant for spatial transcriptomics, where dynamical reconstruction should preserve tissue structure in addition to matching expression patterns. While static formulations have been widely used for such structure-aware alignment, a general dynamic formulation for reconstructing continuous trajectories is still missing. We introduce Travelling Pair Dynamical Alignment and Trajectory Estimation (TP-DATE), a theoretical and computational framework to generalize GW-OT dynamically in a simulation-free manner. We formulate a broad class of static and dynamic Quadratic-form OT (QOT) through path actions and prove the static dynamic equivalence. We further develop travelling-pair flow matching, which allows interacting conditional paths and marginalizes their interactions into a single vector field. On synthetic and real spatial transcriptomics data, TP-DATE better preserves spatial structure and improves continuous 3D dynamics reconstruction.

cs.LG↗

arXiv2026

Structure-preserving diffuse-domain accelerated saddle dynamics for wetting transitions

Wetting transitions on textured substrates play a central role in the design of functional surfaces, but resolving their transition mechanisms requires efficient exploration of complex energy landscapes. In particular, saddle points are essential for revealing the connectivity among metastable states and transition pathways. In this work, we develop a diffuse-domain accelerated saddle dynamics (DD-ASD) framework for mass-constrained wetting transitions on complex textured substrates. The diffuse-domain formulation embeds complex solid geometries into a Cartesian grid, avoiding body-fitted mesh construction in repeated saddle-point searches. An orthogonal projection is applied consistently to the phase-field state and unstable-direction dynamics, preserving the prescribed droplet mass at the discrete level. Momentum acceleration is incorporated into the saddle dynamics to improve the efficiency of high-index saddle searches. Under the stated local spectral and exact-eigenspace assumptions, we establish a local convergence rate of $1-\mathcal{O}(1/\sqrtκ)$ on the effective mass-conserving subspace. Numerical experiments demonstrate the effectiveness of the proposed method in resolving wetting solution landscapes, transition pathways, and energy barriers on textured substrates. We further extend the framework to fully three-dimensional wetting-landscape computations.

math-ph↗

arXiv2026

OREO: Fidelity Alignment in 3D Generation via On-the-fly Rendering-Editing Optimization

Despite recent advancements in 3D generation, models often struggle to produce assets with high visual fidelity. To bridge this gap, we propose OREO, an alignment framework that enhances the realism of 3D generators by leveraging rich 2D diffusion priors. Instead of relying on static datasets, OREO establishes a dynamic optimization loop that produces on-the-fly edited renderings as 2D pseudo-targets. At its core, we introduce Reinforced Editing, which utilizes a 2D model to refine rendered views of the 3D output, enhancing their overall visual fidelity while preserving the underlying geometry, viewpoint, and content. These refined views serve as high-quality supervision targets, enabling the 3D generator to learn from its own generated samples and progressively improve its visual quality. Experiments demonstrate that OREO effectively improves upon pre-trained baselines, producing 3D assets with enhanced visual realism.

cs.CV↗

arXiv2026

Isolated singularities of solutions to the Yamabe equation in higher dimensions I

We complete, in the Riemannian setting, the classification of isolated singularities of positive solutions to the Yamabe equation for $3\le n\le 29$, initiated in the flat case by Caffarelli, Gidas, and Spruck in 1989. Precisely, every isolated singularity is either removable or asymptotic to a Fowler solution. The sharpness of this range is established by counterexamples for $n\ge30$ in our subsequent paper.

math.AP↗

arXiv2026

A Multimodal Large Language Model-Driven Framework for Context-Aware UAV Emergency Landing Site Selection

Safe UAV emergency landing requires more than just identifying flat terrain; it demands understanding complex semantic risks (e.g., crowds, temporary structures) invisible to traditional geometric sensors. In this paper, we propose a novel framework leveraging Remote Sensing (RS) imagery and Multimodal Large Language Models (MLLMs) for global context-aware landing site assessment. Unlike local geometric methods, our approach employs a coarse-to-fine pipeline: first, a lightweight semantic segmentation module efficiently pre-screens candidate areas; second, a vision-language reasoning agent fuses visual features with Point-of-Interest (POI) data to detect subtle hazards. To validate this approach, we construct and release the Emergency Landing Site Selection (ELSS) benchmark at https://github.com/chunlianghua/ELSS-dataset. ELSS includes a remote-sensing subset for simulation validation and a UAV aerial-video subset for real-world testing. Experiments demonstrate that our framework significantly outperforms geometric baselines in risk identification accuracy. Furthermore, qualitative results confirm its ability to generate human-like, interpretable justifications, enhancing trust in automated decision-making.

cs.CV↗

arXiv2026

The Lange-Stuhler Theorem via Gerbes

We recast the Lange-Stuhler theorem as the universal property of a $2$-cartesian square of classifying stacks. This formulation shows that every Frobenius-periodic vector bundle of rank $r$ on an arbitrary fibered category in characteristic $p>0$ is trivialized by a finite étale torsor, without geometric hypotheses or restrictions on the ground field beyond its characteristic. The construction extends from $\mathrm{GL}_n$ to group schemes over finite fields that are connected affine or of finite type.

math.AG↗

arXiv2026

Less Is More in the Long Tail: Stage-Adaptive Sample Selection for Annotation-Efficient Dense Prediction

Deep learning performance generally improves with increasing training data, yet this scaling is fundamentally constrained by annotation cost in large-scale dense prediction tasks with long-tailed category distributions, where pixel- or voxel-level annotation is prohibitively expensive. We propose SASS (Stage-Adaptive Sample Selection), a stage-adaptive data-selection framework for pool-based active learning in long-tailed dense prediction. SASS combines three components: label-free self-supervised gradient scoring, prior-guided category rebalancing with validation-driven feedback, and stage-adaptive acquisition aligned with model training dynamics. This design avoids candidate ground-truth masks during gradient scoring while making acquisition responsive to long-tail imbalance and evolving representations. We evaluate SASS on a multimodal 3D medical segmentation testbed comprising over 100,000 samples spanning 108 anatomical structures. SASS recovers 98.3% of full-dataset performance with a 40% training-pool annotation budget, outperforming BADGE by 5.1 percentage points. Moreover, SASS exhibits a statistically supported less-is-more pattern, surpassing full-dataset training at the Hard-group level and, at the structure level, for the pancreas and gallbladder. More broadly, SASS shows that annotation-efficient learning depends not only on which samples are selected, but also on how the annotation budget is distributed across categories and when model-derived scores begin to guide selection.

cs.CV↗

arXiv2026

Complementary Time- and Distance-Based Methods for Cosmogenic $^{9}$Li/$^{8}$He Background Estimation

Cosmogenic $^{9}$Li and $^{8}$He isotopes constitute an important correlated background in low-energy neutrino experiments because their $β$-delayed neutron decay signatures can mimic inverse beta decay signals. Conventional estimates based on the time since the last muon become challenging at high muon rates, while muon-related vetoes further reduce the residual isotope statistics. We extend the conventional time fit to an event-level muon-categorized joint time (J-MuCAT) fit, which uses the time to the most recent preceding muon in each energy-loss category. A complementary estimate is obtained from the candidate-to-muon-track distance distribution (TraDiTS). The muon-uncorrelated component is determined from the far-distance region and subtracted. This estimate is then used to constrain the J-MuCAT fit, defining the distance-constrained J-MuCAT (DCJ-MuCAT) fit. In detector-level simulation with successive cosmogenic-background vetoes, DCJ-MuCAT reduces the statistical uncertainty by more than $40\%$ relative to J-MuCAT and by more than $10\%$ relative to the TraDiTS. The fitted results remain consistent with the simulation truth. Applicability studies further show good performance over a broad range of muon rates and isotope fractions. The proposed framework provides a practical approach for estimating residual $^{9}$Li/$^{8}$He backgrounds in large neutrino detectors.

hep-ex↗

arXiv2026

Real-Frequency Diagrammatics with Pole Representations

Finite-temperature many-body theories are typically formulated in imaginary time, where they are amenable to efficient numerical treatment. However, the extraction of spectral properties from the imaginary axis is difficult and requires numerical analytic continuation, while a direct formulation on the real axis requires prohibitively expensive numerical quadrature. In this paper, we represent the objects of many-body perturbation theory using systematically improvable pole and moment expansions and develop the corresponding framework for evaluating diagrammatic contributions directly on the real axis. We introduce a controlled recompression procedure that prevents the proliferation of poles under algebraic and diagrammatic operations. We benchmark the approach on paradigmatic quantum impurity problems and apply it to self-consistent calculations of the uniform electron gas. The resulting spectral quantities converge systematically towards numerical accuracy. This work provides a practical framework for controlled and systematically improvable calculations of spectral quantities within diagrammatic many-body theories, including self-consistent conserving approaches.

cond-mat.str-el↗

arXiv2026

SWE-Universe: Scale Real-World Verifiable Environments to Millions

We propose SWE-Universe, a scalable and efficient framework for automatically constructing real-world software engineering (SWE) verifiable environments from GitHub pull requests (PRs). To overcome the prevalent challenges of automatic building, such as low production yield, weak verifiers, and prohibitive cost, our framework utilizes a building agent powered by an efficient custom-trained model. This agent employs iterative self-verification and in-loop hacking detection to ensure the reliable generation of high-fidelity, verifiable tasks. Using this method, we scale the number of real-world multilingual SWE environments to a million scale (807,693). We demonstrate the profound value of our environments through large-scale agentic mid-training and reinforcement learning. Finally, we applied this technique to Qwen3-Max-Thinking and achieved a score of 75.3% on SWE-Bench Verified. Our work provides both a critical resource and a robust methodology to advance the next generation of coding agents.

cs.SE↗

arXiv2026

General Collaborative Intelligence: Architecting Cognition for Resilient Multi-Agent Ecosystems

Multi-agent unmanned systems are moving from isolated, ego-centric sensing toward collaborative intelligence, in which distributed agents exchange compact features to overcome a local observation trap that no single agent can escape: occlusions, finite sensor range, and environmental degradation. The field has matured across architectural, communication, embodied, resilience, and trust dimensions, yet existing surveys examine these dimensions in isolation and rarely expose their dependencies. This review offers a unified synthesis through two complementary lenses. The first is a five-dimensional taxonomy spanning collaboration stage, communication paradigm, fusion architecture, learning strategy, and application domain. The second is three cognitive synergy conditions, Semantic Disambiguation, Pragmatic Information Exchange, and Proactive Informational Foraging, that turn cognitive synergy into operational criteria. Across these lenses we survey collaboration architectures and topologies, neural-communication co-design that treats the channel as a differentiable pipeline component, embodied action-perception loops via multi-agent reinforcement learning, and resilience mechanisms for synchronization, uncertainty quantification, and label-efficient learning. We then map these advances onto four operational domains, V2X, unmanned aerial, industrial logistics, and smart cities, and onto the safety-privacy-utility triad. To counter benchmark saturation and evaluation fragmentation, we propose GCI-Bench, a five-pillar scoring protocol with a maturity model that makes the trade-offs of collaborative methods comparable across studies. A critical reflection on reproducibility, the sim-to-real gulf, and conditions under which collaboration degrades performance identifies open challenges and charts directions toward general collaborative intelligence under real-world uncertainty.

cs.CV↗