Searcharxiv⌕ Search

arXiv subjects

Bohan Zhang

Publications and source records attributed to Bohan Zhang.

At least 19 recordsLinked to original sources
arXiv2026

ORBITALIF: An Efficient Spiking Federated Learning Framework for Onboard Cloud Removal

Low-earth-orbit (LEO) satellites enable high-resolution, large-scale Earth observation for applications such as disaster monitoring and environmental surveillance. However, cloud coverage often obscures the Earth's surface, and conventional cloud-removal pipelines that download cloudy images to ground stations for processing suffer from limited contact windows, constrained satellite-to-ground bandwidth, and high latency. In this work, we propose a novel satellite federated learning framework for cloud removal across LEO constellations, named orbital attention leaky integrate-and-fire (OrbitALIF). OrbitALIF performs both onboard training and inference using a compact 2.30,M-parameter spiking neural network (SNN) backbone with an adaptive gated fusion module (AGFM) and a spectral-spatial hybrid attention module (SHAM), combined with a decentralized federated learning strategy that shares model weights via inter-satellite links. Our experiments show that OrbitALIF achieves competitive cloud removal quality while consuming only 0.287,mJ per inference on neuromorphic hardware, a 72.3 times (98.6%) energy reduction versus an equivalent artificial neural network (ANN).

cs.NE↗
arXiv2026

Weightless Fine-Tuning: Personalizing LLMs via Logit-Space Transport

Supervised fine-tuning (SFT) is a standard approach for adapting LLMs to a target distribution, but in settings such as personalization, where each author requires separate weight access, optimization, storage, and retraining, its costs become prohibitive. We propose Weightless Fine-Tuning (WFT), a training-free decoding-time method that approximates the distributional effect of SFT without weight updates. WFT computes supervised residuals on an author's training sequence and transports them to the current prompt through a cross-prefix transport operator estimated from dropout-induced cross-covariance. The operator captures how a perturbation at one context propagates to predictions at another, replacing gradient-based parameter updates with logit-space corrections. On three LaMP personalization benchmarks, WFT achieves the best average performance across datasets, matches or exceeds SFT on individual tasks, and outperforms other lightweight baselines on average. In a budget-controlled comparison, WFT approaches SFT performance using less than 7% of the effective computation. Logit-level analysis shows a cosine similarity of 0.875 between the logit shifts induced by WFT and SFT over 95% of the next-token probability mass, suggesting that WFT captures the distributional effect of supervised adaptation without modifying model weights.

cs.LG↗
arXiv2026

A Wearable Stiffness-Rendering Haptic Device with a Honeycomb Jamming Mechanism for Bilateral Teleoperation

This paper addresses the challenge of providing kinesthetic feedback in bilateral teleoperation by designing a wearable, lightweight (20 g), and compact haptic device, the HJ-Haptic, utilizing a honeycomb jamming mechanism for object stiffness rendering. The HJ-Haptic device can vary its stiffness, from 1.15 N/mm to 2.64 N/mm, using a 30 kPa vacuum pressure. We demonstrate its implementation in a teleoperation framework, enabling operators to adjust grip force based on a reliable haptic feedback on object stiffness. A three-point flexural test on the honeycomb jamming mechanism and teleoperated object-grasping tasks were conducted to evaluate the device's functionality. Our experiments demonstrated a small RMSE and strong correlations in teleoperated motion, stiffness rendering, and interaction force feedback. The HJ-Haptic effectively adjusts its stiffness in response to real-time gripper feedback, mimicking the sensation of direct object grasping with hands. The device's use of vacuum pressure ensures operator safety by preventing dangerous outcomes in case of gas leakage or material failure. Incorporating the HJ-Haptic into the teleoperation framework provided the reliable perception of object stiffness and stable teleoperation. This study highlights the potential of the honeycomb jamming mechanism for enhancing haptic feedback in various applications, including teleoperation scenarios, as well as interactions with extended-reality environments.

cs.RO↗
arXiv2026

Consistent Feature Transport for Image Relighting

Image relighting modifies illumination while preserving non-lighting content such as identity and geometry. Existing diffusion-based methods often suffer from unstable illumination changes or inconsistent content preservation under complex lighting, as they lack an explicit mechanism to learn feature transformations between images. We reformulate relighting as an illumination feature transport problem and introduce Consistent Feature Transport (CFT), a training principle that explicitly enforces illumination-consistent transport between source and target image distributions. Built upon rectified flow, CFT jointly models noise-to-image generation and illumination-consistent source-to-target transport through trajectory-level supervision. This dual-transport formulation encourages isolation of illumination-specific variations while preserving content-aligned features. To support complex lighting scenarios, we construct a large-scale portrait relighting dataset with diverse relighting effects. Experiments show consistent improvements over existing state-of-the-art relighting approaches and demonstrate that CFT can generalize to other editing tasks, including style transfer. Code is available at https://github.com/Dixin-Lab/CFT.

cs.CV↗
arXiv2026

Bifurcation reordering programs snap-through symmetry in folded elastic ribbons

Snap-through in slender elastic structures is often viewed as a sudden transition between stable configurations, yet the pathway taken during this transition can differ fundamentally. A structure may snap while preserving symmetry, or first lose symmetry and pass through an asymmetric state before reaching its final configuration. What selects between these pathways remains less well understood, especially when the geometry and loading are themselves symmetric. Here, we show that snap-through symmetry can be programmed by reordering competing bifurcations in folded elastic ribbons. We study an elastic ribbon with two localized folds placed symmetrically about the midpoint and show that varying the fold position changes the relative order of two instabilities: a symmetry-breaking pitchfork bifurcation and a saddle-node bifurcation on the symmetry-preserving branch. When the pitchfork bifurcation occurs first, the ribbon loses symmetry before snapping and follows an asymmetric pathway. Conversely, when the saddle-node bifurcation occurs first, the ribbon loses stability while remaining on the symmetric branch, resulting in a symmetry-preserving transition. Combining experiments, discrete differential geometry simulations and numerical continuation, we map this exchange in bifurcation ordering and construct a phase diagram that predicts the switch between asymmetric and symmetric snap-through regimes. A reduced-order double-mass von Mises truss model captures the same mechanism as a generic competition between symmetry-breaking and symmetry-preserving instabilities. These results establish bifurcation reordering as a geometric mechanism for programming snap-through pathways in slender elastic structures, offering a design principle for multistable systems, morphing structures and instability-based mechanical devices.

cond-mat.soft↗
arXiv2026

SpreadsheetBench 2: Evaluating Agents on End-to-End Business Spreadsheet Workflows

Spreadsheets are widely used for business analysis, financial modeling, reporting, and decision-making. However, most existing spreadsheet benchmarks evaluate isolated operations such as single-formula generation or local cell edits, and therefore fail to capture end-to-end workflows in realistic business settings. We introduce \textsc{SpreadsheetBench 2}, a workflow-level benchmark for spreadsheet agents that covers three task categories: generation, debugging, and visualization. The benchmark is constructed from authentic business data, including financial reports and corporate filings, and is annotated and validated by domain experts. The benchmark contains 321 tasks; each instance averages 11.8 worksheets and requires 593.5 cell modifications, reflecting large multi-sheet workbooks with cross-sheet dependencies. We evaluate eight frontier large language models under a unified multi-turn agent scaffold, and additionally include several LLM-based spreadsheet products as complementary baselines. Results show that current systems remain far from reliable on real-world workflows: the best model achieves 34.89\% overall task accuracy, and debugging accuracy is as low as 12.00\%. Trajectory analysis and a failure taxonomy further indicate that insufficient spreadsheet inspection and incorrect target-cell selection are the dominant bottlenecks. Together, these findings position \textsc{SpreadsheetBench 2} as a challenging testbed for advancing reliable spreadsheet automation. Project page: https://spreadsheetbench.github.io/

cs.SE↗
arXiv2026

LongLive-2.0: An NVFP4 Parallel Infrastructure for Long Video Generation

We present LongLive-2.0, an NVFP4-based parallel infrastructure throughout the full training and inference workflow of long video generation, addressing speed and memory bottlenecks. For training, we introduce sequence-parallel autoregressive (AR) training, instantiated as Balanced SP, which co-designs the efficient teacher-forcing layout with SP execution by pairing clean-history and noisy-target temporal chunks on each rank, enabling a natural teacher-forcing mask with SP-aware chunked VAE encoding. Combined with NVFP4 precision, it reduces GPU memory cost and accelerates GEMM computation during training, the proportion of which increases as video length grows. Moreover, we show that a high-quality infrastructure and dataset enable a remarkably clean training pipeline. Unlike existing Self-Forcing series methods that rely on ODE initialization and subsequent distribution matching distillation (DMD), LongLive-2.0 directly tunes a diffusion model into a long, multi-shot, interactive auto-regressive (AR) diffusion model. It can be further converted to real-time generation (4 to 2 denoising steps) with standalone LoRA weights. For inference on Blackwell GPUs, we enable W4A4 NVFP4 inference, quantize KV cache into NVFP4 for memory savings, and boost end-to-end throughput with asynchronous streaming VAE decoding. On non-Blackwell GPU architectures, we deploy SP inference to match the speed on Blackwell GPUs, while the quantized KV cache can lower inter-GPU communication of SP. Experiments show up to 2.15x speedup in training, and 1.84x in inference. LongLive-2.0-5B achieves 45.7 FPS inference while attaining strong performance on benchmarks. To our knowledge, LongLive-2.0 is the first NVFP4 training and inference system for long video generation.

cs.CV↗
arXiv2026

Convergence Analysis of Hessian-Damped Tikhonov Regularized Dynamics with Oscillation Control for Convex-Concave Bilinear Saddle Point Problems

In this paper, we propose a class of general second-order primal-dual dynamical systems with Tikhonov regularization and Hessian-driven damping for solving convex-concave bilinear saddle point problems. The proposed dynamical system incorporates five general time-varying terms: viscous damping, time scaling, extrapolation, Tikhonov regularization, and Hessian-driven damping parameters. Under suitable parametric conditions, we analyze the asymptotic convergence properties of the dynamical system by constructing appropriate Lyapunov functions. Specifically, we obtain the convergence rate of the primal-dual gap and the boundedness of trajectories in the proposed dynamical system, and provide some integral estimates. Furthermore, we theoretically prove that the trajectories generated by the dynamical system converge strongly to the minimum-norm solution of the saddle point problem, and fully demonstrate that Hessian-driven damping can effectively alleviate oscillations. Finally, numerical experiments are conducted to verify the validity of the above theoretical results.

math.OC↗
arXiv2026

PACO: Proxy-Task Alignment and Online Calibration for On-the-Fly Category Discovery

On-the-Fly Category Discovery (OCD) requires a model, trained on an offline support set, to recognize known classes while discovering new ones from an online streaming sequence. Existing methods focus heavily on offline training. They aim to learn discriminative representations on the support set so that novel classes can be separated at test time. However, their discovery mechanism at inference is typically reduced to a single threshold. We argue that this paradigm is fundamentally flawed as OCD is not a static classification problem, but a dynamic process. The model must continuously decide 1) whether a sample belongs to a known class, 2) matches an existing novel category, or 3) should initiate a new one. Moreover, prior methods treat the support set as fixed knowledge. They do not update their decision boundaries as new evidence arrives during inference. This leads to unstable and inconsistent category formation. Our experiments confirm these issues. With properly calibrated and adaptive thresholds, substantial improvements can be achieved, even without changing the representation. Motivated by this, we propose PACO, a support-set-calibrated, tree-structured online decision framework. The framework models inference as a sequence of hierarchical decisions, including known-class routing, birth-aware novel assignment, and attach-versus-create operations over a dynamic prototype memory. Furthermore, we simulate the proxy discovery process to initialize the thresholds during offline training to align with inference. Thresholds are continuously updated during inference using mature novel prototypes. Importantly, PACO requires no heavy training and no dataset-specific tuning. It can be directly integrated into existing OCD pipelines as an inference-time module. Extensive experiments show significant improvements over SOTA baselines across seven benchmarks.

cs.CV↗
arXiv2026

AutonomyLens: A Self-Evolving Simulation-Based Testing Loop for Autonomous Systems

Software engineering practices for validating autonomous cyber-physical systems (e.g., Uncrewed Aerial Vehicles) remain fragmented across scenario design, simulation execution, and telemetry analysis, limiting traceability between requirements, tests, and evidence. This fragmentation reduces reproducibility, slows debugging and iteration, and hinders systematic assurance under complex and evolving environmental conditions. We present AutonomyLens, an LLM-driven framework that integrates scenario specification, simulation execution, and telemetry analysis into a unified validation workflow. AutonomyLens enables developers to translate high-level validation intent into executable, temporally evolving scenarios, automatically run simulations, and perform context-aware analysis of resulting system behavior. The framework introduces (i) a structured representation for mission-level scenarios, (ii) an automated execution pipeline, (iii) analysis mechanisms that align telemetry with scenario context to produce actionable insights, and (iv) counterfactual scenario generation that closes the loop by refining and synthesizing new test cases from observed failures. We describe the early-stage design of AutonomyLens, discuss key challenges in building integrated validation workflows for autonomous systems, and outline how such an approach can improve traceability, reproducibility, and scalability in autonomy validation.

cs.SE↗
arXiv2026

Spin-adapted neural network backflow for symmetry-preserving simulations of strongly correlated electrons

Strongly correlated molecules often contain dense manifolds of low-lying spin states, making total-spin symmetry essential for predictive electronic-structure theory. Neural-network quantum states provide flexible variational wavefunctions, but commonly used fermionic architectures do not enforce this symmetry and can therefore converge to spin-contaminated states with misleading energies and properties. Here we introduce a spin-adapted neural-network backflow (SA-NNBF) ansatz in second quantization, which combines configuration-dependent spatial orbitals with a compressed spin eigenfunction. A projected tensor compression scheme for spin eigenfunctions and a particle-hole representation make variational Monte Carlo calculations with SA-NNBF practical for active spaces containing more than one hundred electrons. Across hydrogen chains and iron-sulfur clusters, SA-NNBF eliminates spin contamination and consistently achieves lower variational energies than standard NNBF with a comparable number of parameters. For the CAS(113e,76o) active-space model of FeMoco, SA-NNBF yields a highly compact spin-adapted variational state, achieving an energy competitive with recent spin-adapted DMRG calculations at bond dimension $D=10000$ while using orders of magnitude fewer parameters. Our work establishes a general framework for developing spin-symmetry-preserving neural-network quantum states for chemically realistic strongly correlated electrons.

physics.chem-ph↗
arXiv2026

From Folding Mechanics to Robotic Function: A Unified Modeling Framework for Compliant Origami

Origami inspired architectures offer a powerful route toward lightweight, reconfigurable, and programmable robotic systems. Yet, a unified mechanics framework capable of seamlessly bridging rigid folding, elastic deformation, and stability driven transitions in compliant origami remains lacking. Here, we introduce a geometry consistent modeling framework based on discrete differential geometry (DDG) that unifies panel elasticity and crease rotation within a single variational formulation. By embedding crease panel coupling directly into a mid edge geometric discretization, the framework naturally captures rigid folding limits, distributed bending, multistability, and nonlinear dynamic snap through within one mechanically consistent structure. This unified description enables programmable control of stability and deformation across rigid and compliant regimes, allowing origami structures to transition from static folding mechanisms to active robotic modules. An implicit dynamic formulation incorporating gravity, contact, friction, and magnetic actuation further supports strongly coupled multiphysics simulations. Through representative examples spanning single fold bifurcation, deployable Miura membranes, bistable Waterbomb modules, and Kresling based crawling robots, we demonstrate how geometry driven mechanics directly informs robotic functionality. This work establishes discrete differential geometry as a foundational design language for intelligent origami robotics, enabling predictive modeling, stability programming, and mechanics guided robotic actuation within a unified computational platform.

cs.RO↗
arXiv2026

Learning through Creation: A Hash-Free Framework for On-the-Fly Category Discovery

On-the-Fly Category Discovery (OCD) aims to recognize known classes while simultaneously discovering emerging novel categories during inference, using supervision only from known classes during offline training. Existing approaches rely either on fixed label supervision or on diffusion-based augmentations to enhance the backbone, yet none of them explicitly train the model to perform the discovery task required at test time. It is fundamentally unreasonable to expect a model optimized on limited labeled data to carry out a qualitatively different discovery objective during inference. This mismatch creates a clear optimization misalignment between the offline learning stage and the online discovery stage. In addition, prior methods often depend on hash-based encodings or severe feature compression, which further limits representational capacity. To address these issues, we propose Learning through Creation (LTC), a fully feature-based and hash-free framework that injects novel-category awareness directly into offline learning. At its core is a lightweight, online pseudo-unknown generator driven by kernel-energy minimization and entropy maximization (MKEE). Unlike previous methods that generate synthetic samples once before training, our generator evolves jointly with the model dynamics and synthesizes pseudo-novel instances on the fly at negligible cost. These samples are incorporated through a dual max-margin objective with adaptive thresholding, strengthening the model's ability to delineate and detect unknown regions through explicit creation. Extensive experiments across seven benchmarks show that LTC consistently outperforms prior work, achieving improvements ranging from 1.5 percent to 13.1 percent in all-class accuracy. The code is available at https://github.com/brandinzhang/LTC

cs.CV↗
arXiv2026

CoSA: Compressed Sensing-Based Adaptation of Large Language Models

Parameter-Efficient Fine-Tuning (PEFT) has emerged as a practical paradigm for adapting large language models (LLMs) without updating all parameters. Most existing approaches, such as LoRA and PiSSA, rely on low-rank decompositions of weight updates. However, the low-rank assumption may restrict expressivity, particularly in task-specific adaptation scenarios where singular values are distributed relatively uniformly. To address this limitation, we propose CoSA (Compressed Sensing-Based Adaptation), a new PEFT method extended from compressed sensing theory. Instead of constraining weight updates to a low-rank subspace, CoSA expresses them through fixed random projection matrices and a compact learnable core. We provide a formal theoretical analysis of CoSA as a synthesis process, proving that weight updates can be compactly encoded into a low-dimensional space and mapped back through random projections. Extensive experimental results show that CoSA provides a principled perspective for efficient and expressive multi-scale model adaptation. Specifically, we evaluate CoSA on 10 diverse tasks, including natural language understanding and generation, employing 5 models of different scales from RoBERTa, Llama, and Qwen families. Across these settings, CoSA consistently matches or outperforms state-of-the-art PEFT methods.

cs.LG↗
arXiv2026

ContextBench: A Benchmark for Context Retrieval in Coding Agents

LLM-based coding agents have shown strong performance on automated issue resolution benchmarks, yet existing evaluations largely focus on final task success, providing limited insight into how agents retrieve and use code context during problem solving. We introduce ContextBench, a process-oriented evaluation of context retrieval in coding agents. ContextBench consists of 1,136 issue-resolution tasks from 66 repositories across eight programming languages, each augmented with human-annotated gold contexts. We further implement an automated evaluation framework that tracks agent trajectories and measures context recall, precision, and efficiency throughout issue resolution. Using ContextBench, we evaluate four frontier LLMs and five coding agents. Our results show that sophisticated agent scaffolding yields only marginal gains in context retrieval ("The Bitter Lesson" of coding agents), LLMs consistently favor recall over precision, and substantial gaps exist between explored and utilized context. ContextBench augments existing end-to-end benchmarks with intermediate gold-context metrics that unbox the issue-resolution process. These contexts offer valuable intermediate signals for guiding LLM reasoning in software tasks.

cs.LG↗
arXiv2026

Tikhonov-Regularized Second-Order Primal-Dual Dynamics for Convex-Concave Bilinear Saddle Point Problems

This paper develops a Lyapunov-based coefficient-compatible framework for a class of second-order primal-dual dynamical systems with Tikhonov regularization. Unlike existing accelerated inertial primal-dual dynamics, whose analyses are typically carried out under prescribed damping profiles, the proposed framework characterizes the interplay among the viscous damping coefficient, the time-scaling coefficient, the linear extrapolation parameter, and the Tikhonov regularization parameter through differential compatibility conditions, thereby accommodating time-varying coefficients beyond prescribed damping profiles. Within this framework, we investigate two cases in which the Tikhonov regularization parameter decays to zero at different rates. When the Tikhonov regularization parameter decays to zero relatively rapidly, the fast convergence rate of the primal-dual gap is preserved, whereas a slower decay rate yields refined asymptotic convergence estimates. We further show that a sufficiently slow decay ensures strong convergence of the trajectories to the minimum-norm saddle point. To establish the latter property, we introduce a moving-saddle Lyapunov framework centered at the instantaneous regularized saddle point and derive verifiable coefficient conditions ensuring the required tracking behavior. Finally, three numerical examples are presented to illustrate the theoretical results.

math.OC↗
arXiv2026

Sparse-RL: Breaking the Memory Wall in LLM Reinforcement Learning via Stable Sparse Rollouts

Reinforcement Learning (RL) has become essential for eliciting complex reasoning capabilities in Large Language Models (LLMs). However, the substantial memory overhead of storing Key-Value (KV) caches during long-horizon rollouts acts as a critical bottleneck, often prohibiting efficient training on limited hardware. While existing KV compression techniques offer a remedy for inference, directly applying them to RL training induces a severe policy mismatch, leading to catastrophic performance collapse. To address this, we introduce Sparse-RL empowers stable RL training under sparse rollouts. We show that instability arises from a fundamental policy mismatch among the dense old policy, the sparse sampler policy, and the learner policy. To mitigate this issue, Sparse-RL incorporates Sparsity-Aware Rejection Sampling and Importance-based Reweighting to correct the off-policy bias introduced by compression-induced information loss. Experimental results show that Sparse-RL reduces rollout overhead compared to dense baselines while preserving the performance. Furthermore, Sparse-RL inherently implements sparsity-aware training, significantly enhancing model robustness during sparse inference deployment. The corresponding training data and code are publicly available on the repository.

cs.LG↗
arXiv2026

Who Owns the Text? Design Patterns for Preserving Authorship in AI-Assisted Writing

AI writing assistants can reduce effort and improve fluency, but they may also weaken writers' sense of authorship. We study this tension with an ownership-aware co-writing editor that offers on-demand, sentence-level suggestions and tests two common design choices: persona-based coaching and style personalization. In an online study (N=176), participants completed three professional writing tasks: an email without AI help, a proposal with generic AI suggestions, and a cover letter with persona-based coaching, while half received suggestions tailored to a brief sample of their prior writing. Across the two AI-assisted tasks, psychological ownership dropped relative to unassisted writing (about 0.85-1.0 points on a 7-point scale), even as cognitive load decreased (about 0.9 points) and quality ratings stayed broadly similar overall. Persona coaching did not prevent the ownership decline. Style personalization partially restored ownership (about +0.43) and increased AI incorporation in text (+5 percentage points). We distill five design patterns: on-demand initiation, micro-suggestions, voice anchoring, audience scaffolds, and point-of-decision provenance, to guide authorship-preserving writing tools.

cs.HC↗