arXiv subjects

Zihan Yang

Publications and source records attributed to Zihan Yang.

At least 19 recordsLinked to original sources

arXiv2026

Agentic Visual Generation: From Generative Models to Agentic Control

Visual generation is evolving from generative models used through a single invocation into agentic control processes that can plan, select tools, inspect intermediate synthesized outputs, revise failures, and reuse prior experience. In most existing systems, the controller is an LLM or VLM, while visual generation models serve as tools or executors. However, existing work lacks a consistent criterion for determining when a generation system becomes agentic. Planning depth, tool use, multi-role collaboration, and reinforcement learning are often treated as evidence of agenticity, even though none of them necessarily determines which generation decisions the controller can make. We organize the field according to what the controller can directly control in the generation process. At L1 Conditioning Control, the controller prepares the input to a predetermined generator but does not control which visual operation is executed. At L2 Execution Control, it selects and invokes actual generation, editing, rendering, or other content-modifying operations. At L3 Outcome-Adaptive Control, it observes an intermediate outcome and uses that observation to change a subsequent operation within the current task. At L4 Experience-Adaptive Control, it retains experience from completed tasks and uses that experience to change decisions on future tasks. L0 Fixed Support separately denotes generators, editors, evaluators, reward models, benchmarks, and fixed pipelines without a deployed controller that makes generation-level decisions. These levels describe a progressively broader decision-making scope rather than model size, system complexity, output quality, tool or role count, or training method. Applying this framework across image, video, editing, 3D, world, slide, and user-interface generation reveals how controller capabilities have evolved and how their mechanisms are distributed across levels.

cs.CV↗

arXiv2026

OCGQuant: Outlier-Companion Grouping for NVFP4 Quantization

NVFP4 is an efficient microscaling format for low-bit inference, but activation outliers can still degrade quantization accuracy within NVFP4 blocks. Within each quantization block, large activations can dominate the block scale, increasing the quantization error of the remaining values sharing the same scale. Existing post-training quantization (PTQ) methods mitigate outlier errors through strategies such as mixed precision, rotation, or residual compensation, but these approaches are either not specifically tailored to NVFP4 or introduce additional computation. In this work, we revisit NVFP4 from a channel-grouping perspective and define the reducible error incurred by remaining block values under the scale set by the block maximum as Collateral Quantization Error. Based on this insight, we propose OCGQuant, a post-training quantization method centered on Outlier-Companion Grouping (OCG), which adaptively pairs outlier channels with low-magnitude companion channels to improve NVFP4 activation block composition. Experiments on Llama3 and Qwen3 show that OCGQuant achieves the lowest WikiText-2 perplexity and highest average downstream accuracy among evaluated PTQ methods, while maintaining prefill speedup close to RTN and matching its peak decoding memory. Code is available at https://github.com/Eshamont/OCGQuant.

cs.CL↗

arXiv2026

CARO: Contact-Agnostic Residual Observation for Zero-Shot Robust Quadruped Locomotion

We propose CARO, a contact-agnostic residual observation framework for policy adaptation. CARO embeds a fixed-base Euler--Lagrange model into the reinforcement learning control loop and constructs a torque-level residual observation without requiring torque sensors, explicit contact estimation, or vision-based measurements of the floating-base position and linear velocity. A disturbance observer extracts a structured signal representing dynamics mismatch, while the policy learns to exploit this feedback for online adaptation. CARO is trained under the same terrain, command, and domain-randomization conditions as the nominal policy, without specialized disturbance curricula or additional adaptation supervision. Nevertheless, it achieves substantially improved zero-shot robustness in simulation and sim-to-real transfer tasks involving out-of-distribution payloads, center-of-mass shifts, terrain geometries, abrupt dynamics changes, and elevated-platform landings.

cs.RO↗

arXiv2026

Physics Filtering Favors the Generalization of Robot Learning

Living organisms exhibit extraordinary adaptability to unseen environments through their intrinsic physical structures and lifelong feedback-driven learning. Endowing robots with comparable generalization is critical for reliable operation in the real world. While recent approaches attempt to improve generalization by scaling training data, such strategies remain impractical for robotics, where collecting real-world demonstrations at the scale of large language models is prohibitively costly and slow. Contrary to this reliance on massive datasets, we show that robots can generalize effectively under dynamics uncertainties even with limited training data by leveraging a feedback mechanism, namely PhyFilter, that corrects learning outputs with physics-filtered learning residuals. PhyFilter operates as a lightweight, model-agnostic module whose parameters can be automatically optimized through an auto-learning algorithm, eliminating manual tuning and enabling seamless integration with diverse robot policies. We validate PhyFilter across four representative robotic systems, demonstrating that it enables quadruped robots to generalize to unseen terrains, payload variations, and speed ranges; drones to flight under unseen wind disturbances; aerial manipulators to achieve centimeter-level in-air capture despite wind and mass uncertainties; and acceleration differentiators to remain robust with distribution shift. These results show that physics-filtered feedback can serve as a powerful alternative to massive data scaling.

cs.RO↗

arXiv2026

VA-Judger: Reward Modeling from Human Preference Feedback for Joint Video-Audio Generation

Using reinforcement learning to post-train joint video-audio generation models requires a reward signal. Existing methods construct this reward by combining metrics for individual quality dimensions, including audio quality, visual fidelity, and synchronization. However, these metrics evaluate perceptual dimensions separately and fail to capture the overall semantic and temporal coherence among the text prompt, video, and audio that shapes human preferences. Optimizing models against these metrics encourages reward hacking, generating video-audio content that achieves high scores on these metrics yet appears incoherent or unfaithful to human viewers. To address this problem, we first construct a large-scale human-preference dataset VAPref-10K for joint video-audio generation, comprising 9K prompts and 10.3K fine-grained paired comparisons from open-source generation models. We also introduce the VA-Judger-Bench benchmark with both in-domain and out-of-domain model comparisons to evaluate whether reward models truly align with human preferences. We further propose VA-Judger, a chain-of-thought omni-reward model for joint video-audio generation. In particular, VA-Judger first learns from pairs with clear quality gaps to establish structured output and coarse preference discrimination, then distills reliable preference explanations for harder near-quality comparisons via rejection sampling verified against human annotations, and finally performs dimension-wise reinforcement learning that decomposes human feedback into individual quality dimensions for denser reward signals than a single binary preference label. Experiments show that VA-Judger outperforms metric baselines in predicting human preferences on both in-domain and out-of-domain evaluations. Using its human-aligned rewards for post-training audio-video generation model also yields significant improvements in generation quality.

cs.CV↗

arXiv2026

VTInstructor: Visual Trajectory Prompting for Navigation Instruction Generation in Continuous Environments

Navigation instruction generation from ego-centric RGB video in continuous environments is an important yet challenging task for human-robot interaction and scalable dataset construction. Prior instruction generators assume discrete viewpoint graphs with panoramic observations, where trajectory structure is explicit; in continuous environments, however, the agent receives only a dense RGB stream, making trajectory cues difficult to recover. We propose VTInstructor, the first VLN instruction generation framework for continuous environments. Our key idea is to convert implicit trajectory geometry into explicit visual trajectory prompts: EDTC condenses long RGB trajectories into navigation-critical keyframes, VTP overlays path, turn, and goal cues onto these anchors, VTMod injects the resulting trajectory signals into the visual encoder, and VT-GRPO further calibrates this spatial injection during training, all without requiring a navigation graph, pre-built map, or scene reconstruction. On the challenging R2R-CE and RxR-CE Val Unseen benchmarks, VTInstructor sets a new state of the art across all standard NLG metrics, surpassing the strongest baseline by +0.357 CIDEr and +0.109 CIDEr, respectively. Beyond automatic metrics, VTInstructor-generated instructions raise a frozen follower's success rate to 63.3%, a +14.7 percentage-point gain over the best competing instruction source, and provide consistent data augmentation gains of +3 SR points on downstream navigation tasks.

cs.RO↗

arXiv2026

InterPruner: Interactive Structured Pruning via Taylor-Implicit Criterion and Language-Prior Modulator for Multimodal Object Detection

Multimodal object detection proves effective in remote sensing, especially the RGB-Infrared paradigm. The parallel feature extractors provide rich multimodal information for robust detection, yet introduce substantial channel redundancy and computational overhead. Existing pruning methods can reduce channel redundancy, but they are designed for unimodal backbones, overlooking cross-modal interactions and dynamic scene-wise redundancy. In this paper, we propose InterPruner, the first interactive structured channel pruning framework for RGB-infrared object detectors. Specifically, we first derive a Taylor-Implicit Criterion(TIC) to quantify channel importance via high-order Taylor expansion and the implicit function theorem. Then, a Modality Interaction Redundancy Analyzer (MIRA) identifies redundant channels via mutual compensability assessment. Finally, a Scene-Prior Channel Anchor (SPCA) uses language priors as semantic anchors to measure channel-scene relevance for dynamic channel importance estimation. Cross-modality channel pruning for RGB-Infrared detection is yet unexplored. Extensive experiments on RGB-infrared object detection dataset demonstrate that InterPruner maintains high performance with negligible degradation. Specifically, it even achieves a 0.6% mAP increase on the FLIR dataset when pruning 50% of the channels. Code will be available on GitHub to facilitate future work.

cs.CV↗

arXiv2026

NeuroInspector: A Local-First Environment for Inspecting and Annotating Hierarchical Neuroscience Datasets

The growing scale and structural complexity of neuroscience datasets have made dataset inspection an increasingly distinct stage of the research workflow. Existing inspection workflows, however, remain fragmented, often relying on exploratory scripts, manual documentation, and repeated navigation of unfamiliar file structures before meaningful scientific analysis can begin. Here we present NeuroInspector, a lightweight, browser-based environment for inspecting HDF5 and NWB files. Using WebAssembly-based HDF5 parsing (h5wasm), NeuroInspector runs entirely client-side: files are opened directly from local disk, and the application implements no file-upload endpoint, including in its hosted deployment. The tool combines structural navigation, metadata inspection, sampled data previews, and path-level annotation into portable, fingerprinted "project packs" that preserve inspection decisions without modifying the original file. Rather than functioning as an analysis or validation platform, NeuroInspector provides a dedicated, traceable environment for the inspection stage that precedes formal analysis.

q-bio.QM↗

arXiv2026

DyFrDet: Towards Accurate Small Object Detection via Dynamic Frequency Suppression with Label Disambiguation

Despite the remarkable progress over the past decades, accurately identifying small objects remains challenging because of their insufficient visual cues. Previous works typically attempt to construct discriminative representation of the small objects. However, the wide range frequency domain noises and label ambiguities have been greatly overlooked, which significantly hinders the accurate localization. To address these issues, we propose a novel small object detection (SOD) detector termed DyFrDet, which is able to precisely localize the small object by dynamically suppressing the background distractions in frequency domain. Specifically, we propose a Dynamic Frequency-aware Feature Pyramid Network (DyFrFPN) to adaptively suppress low-frequency redundancy and excessive high-frequency noises. The DyFrFPN transforms the hierarchical features into frequency domain representation, and introduces a Dynamic Band Predictor (DBP) to preserve the discriminative components for small object identification. Afterwards, we present a novel Label Disambiguation Module (LDM), which leverages probabilistic distributions to explicitly model and alleviate the inherent ambiguity of target labels, yielding efficient improvement in localization precision of the small objects with low-resolution. Extensive experiments demonstrate that DyFrDet achieves state-of-the-art performance across multiple benchmarks, indicating its effectiveness and robustness in various challenging scenarios. Our code is available at https://github.com/ManOfStory/DyFrDet.

cs.CV↗

arXiv2026

Adaptive Spectrum-Aware Feature Disentangled Network for Small Object Detection

Small Object Detection (SOD) is a fundamental yet challenging problem in computer vision due to its limited spatial resolution and weak visual cues. Although recent approaches have achieved remarkable advances, the background distractors in different frequency spectra still degrade the performance. In this paper, we propose a novel small object detection framework termed SFDNet, which is capable of detecting small objects via efficient spectrum-aware feature disentanglement. Specifically, we propose an Adaptive Spectrum Disentanglement (ASD) module that decomposes backbone features into multiple complementary spectral components, aiming to construct discriminative object-relevant representations by discarding the background distractors for each component. Afterwards, to strengthen the semantic consistency of the similar objects in the same class, we propose a Class-Wise Prototype Distillation (CPD) procedure, which establishes class prototypes for the object instances and enforces the compact representation by efficient prototype distillation. Extensive experiments on multiple challenging benchmarks show that SFDNet outperforms existing state-of-the-art methods by a large margin. Our code is available at https://github.com/ManOfStory/SFDNet.

cs.CV↗

arXiv2026

Staying In Character: Perspective-Bounded Memory For Book-Based Role-Playing Agents

Recent LLM role-playing systems build character agents from novels by extracting characters, scenes, and relations. Yet long-narrative role-playing suffers from two failures: Factual Overreach, where shared retrieval or parametric memory lets a character use facts outside its perspective, and Stylistic Monotony, where profile descriptions flatten a character into a fixed voice. To address these failures, we propose REVERIEMEM, a three-layer memory architecture for book-based character agents. The episodic layer stores first-person scene memories; the semantic layer stores visibility-tagged facts; and the personality layer stores situation-dependent speech and behaviour patterns. For evaluation, we construct KBF-QA, a 4,386-question benchmark over eight novels for testing knowledge boundaries. REVERIEMEM improves Knowledge Boundary Fidelity by 34.6 percentage points over the strongest prior method. On BOOKWORLD's five-dimension pairwise narrative protocol, REVERIEMEM achieves a ~ 79% win rate, suggesting that perspective-bounded memory improves both boundary fidelity and character-grounded narrative generation.

cs.CL↗

arXiv2026

Playful Agentic Robot Learning

Current agentic robot systems can write executable Code-as-Policy programs, observe feedback, and revise behavior across multiple attempts, but they remain largely task-driven: reusable skills are acquired only after explicit instructions. We study Playful Agentic Robot Learning, where an embodied coding agent uses self-directed play as a continual skill-learning stage before downstream tasks arrive. We introduce RATs, Robotics Agent Teams designed for play-time skill acquisition. During play, RATs proposes novel yet learnable exploratory tasks, plans and executes robot-code policies, verifies intermediate progress, diagnoses failures, retries with dense, step-level feedback, and distills successful executions into a persistent code skill library. At test time, the agent reuses relevant skills from this frozen library to help solve new tasks. Experiments in LIBERO-PRO and MolmoSpaces show that play-learned skills improve held-out downstream tasks over no-play and random-play baselines, with 20.6 and 17.0 percentage-point gains over CaP-Agent0 on LIBERO-PRO and MolmoSpaces, respectively. Moreover, the learned skills can be plugged into other inference-time Code-as-Policy agents by simply retrieving them into the context, improving RoboSuite and real-world transfer by 8.9 and 8.8 points, respectively, without finetuning the underlying model.

cs.RO↗

arXiv2026

Carbon Layer Orientation and Closed-Pore Construction Achieving Ultra-Low Specific Surface Area Hard Carbon for High-Performance Na-ion Storage

Addressing the critical trade-off between initial Coulombic efficiency (ICE) and reversible capacity in hard carbon anodes for Na-ion batteries (NIBs), we introduce a novel coupling strategy that combines carbon layer orientation reconstruction with closed-pore construction to produce hard carbon with an ultra-low specific surface area. We demonstrate that the nanographite domains within the hard carbon precursor undergo entropy-driven orientation reconstruction through the synergistic regulation of heteroatom doping and medium-temperature carbonization. This process not only increases interlayer spacing and promotes structural disorder but also enables the formation of dense, closed pores and ultramicropores at domain boundaries via confined atomic migration, while simultaneously encapsulating surface open pores within internal closed ones. Due to this unique pore architecture, our hard carbon exhibits an ultra-low specific surface area of 1.89 m2 g-1 with a markedly higher proportion of closed pores. As a result, our hard carbon achieves a remarkable reversible capacity of 342.3 mAh g-1 at 20 mA g-1, with an exceptional ICE of 90.4% and a dominant plateau capacity of 262.3 mAh g-1 (76.6%) for NIBs. We believe this coupling strategy provides a new paradigm for the structural engineering of high-ICE anode materials in advanced NIBs.

cond-mat.mtrl-sci↗

arXiv2026

Who Is in Mind Matters: Attachment Representations in Early Childhood Synchronize Child-Adult Interacting Brains

Human attachment is distinguished by enduring internalized representations that shapes neurodevelopment and social-emotional functioning. However, as unobservable inner processes mixed with social cues and partner-specific factors, the neurocognitive mechanisms of these representations during real-time interaction remain unclear. Using a novel Remote Partner-Belief Manipulation paradigm in 40 child-mother-stranger trios, we experimentally isolated attachment representations in 3-4-year-olds by manipulating children's partner-belief during remote cooperation. The inner processes were captured from synchrony between partners' EEG, showing that children's mother-partner belief, regardless of the actual partner, significantly enhanced interbrain synchrony. This partner-belief modulation concentrated on children's P4 channel (overlaying the attachment-designated right temporoparietal junction), where synchrony strength correlated to attachment security and children's response acceleration due to mother-partner belief. These findings established attachment representations as an independent, endogenous driver of interbrain synchrony, potentially via children's heightened attention towards their attachment figure, implying the role of symbolic attachment activation when separation.

q-bio.NC↗

arXiv2026

Baton: Explicit Semantic Blueprints for Joint Video-Audio Generation

Current open-source diffusion models struggle to generate stable and synchronized audio-visual content, particularly in scenarios demanding complex semantic reasoning. The root cause is that existing methods rely on coarse text embeddings from off-the-shelf encoders to guide audio-video denoising, which discards fine-grained semantics and, critically, lacks a shared long-horizon plan, leading to uncoordinated denoising trajectories and fragile cross-modal alignment. We propose Baton, the first framework that introduces explicit semantic planning into joint video-audio generation. Our key insight is that complementing coarse text guidance with semantically rich, modality-aware planned tokens, jointly reasoned and mutually aligned before denoising, can simultaneously restore fine-grained semantic detail and establish a shared blueprint that coordinates both audio and video denoising trajectories. Concretely, Baton first introduces the VA-Planner, a multimodal language model equipped with dual semantic alignment towers, where learnable queries cross-attend to both video and audio features to produce a pair of semantically aligned video and audio planned tokens as keyframe-level blueprints. These planned tokens are injected into the diffusion backbone via cross-attention layers, providing temporally grounded guidance complementary to coarse text embeddings. Since planned tokens do not share one-to-one spatial-temporal correspondence with diffusion latents, we further propose Relative Semantic RoPE, a relative positional encoding that maps planned tokens and latents into a shared spatial-temporal coordinate frame, enabling each latent to accurately attend to its positionally corresponding semantic cues. Experiments on benchmarks show the effectiveness of Baton both qualitatively and quantitatively.

cs.CV↗

arXiv2026

Unveiling hidden millihertz quasi-periodic oscillations in 1A 0535+262

Be/X-ray binary pulsars show transient outbursts and complex timing behaviour, including millihertz quasi-periodic oscillations (QPOs), whose physical origin and energy dependence remain poorly understood. We aim to characterise the temporal evolution and energy-dependent properties of the mHz QPO during the 2020 giant outburst of 1A 0535+262. We use the multi-Lorentzian fitting framework to jointly model the power spectra and the real and imaginary parts of the cross-spectrum, incorporating simultaneous broadband X-ray observations from NICER and Insight-HXMT (0.2-120 keV). We report the first detection of weak, but significant, mHz QPOs at low X-ray energies (below 27 keV), extending their detection to a new energy regime. The centroid frequency evolves from 41 to 93 mHz, with the peak root-mean-square (rms) amplitude detected in the 50-65 keV. Throughout the outburst, the QPOs generally exhibit a hard lag between 0.12 pi rad and 0.9 pi rad. However, at the outburst peak, the higher-energy bands (above 35 keV) display a soft lag of up to -0.93 pi rad. We propose that interactions between soft seed photons and an extended outflow located outside the magnetosphere can account for the observed hard lags. Furthermore, we detect a double-peaked mHz QPO only at high energies (E above 35 keV) near peak luminosity. The two peaks maintain an approximately constant separation of 2*nu_spin and exhibit anti-correlated phase evolution. Our results indicate that the mHz QPOs in 1A 0535+262 are closely linked to the coupled evolution of a soft-photon source and a Comptonizing outflow or corona. The joint cross-spectral framework provides a complementary probe of mHz QPOs beyond traditional power-spectral analyses.

astro-ph.HE↗

arXiv2026

Gaussian Process Inference of Stochastic Magneto-Active Dynamics and Viscosity in Swift J1727.8-1613

Linking X-ray variability to the underlying magnetohydrodynamic (MHD) dynamics of black hole X-ray binaries remains challenging. We systematically investigate the stochastic and oscillatory variability of the black hole X-ray binary candidate Swift J1727.8$-$1613 during its 2023 outburst using Gaussian process (GP) regression applied to Insight-HXMT multi-band light curves. The variability is modeled with a physically motivated composite kernel comprising one stochastically driven damped simple harmonic oscillator (SHO) and two damped random walk (DRW) components. The SHO term robustly recovers quasi-periodic oscillations (QPOs) with frequencies $\nu_0 \sim 0.07$--$5$ Hz, consistent with the fundamental Alfv\'en mode of a contracting magnetically confined disk--coronal cavity. The quality factor rises from $Q \sim 3$ to $Q \sim 10$, suggesting increasing coherence of the magnetic cavity. We also find an anti-correlation between QPO frequency and the short DRW damping timescale, supporting our proposed stochastic magneto-active dynamics scenario. Associating the short and long DRW timescales with the local turbulent turnover and thermal adjustment timescales, respectively, we infer an effective viscosity parameter of $\alpha \approx 0.1$, supporting a strongly magnetized accretion flow. Strikingly, near the onset of relativistic jet ejection around MJD 60206, both relaxation timescales collapse toward the 0.1 s sampling limit, suggesting a rapid reorganization of the disk internal energy balance immediately before jet launching. Our results establish GP inference as a powerful route to connecting X-ray timing observables with the dynamical state of black hole accretion flows.

astro-ph.HE↗

arXiv2026

ArcFlow: Unleashing 2-Step Text-to-Image Generation via High-Precision Non-Linear Flow Distillation

Diffusion models have achieved remarkable generation quality, but they suffer from significant inference cost due to their reliance on multiple sequential denoising steps, motivating recent efforts to distill this inference process into a few-step regime. However, existing distillation methods typically approximate the teacher trajectory by using linear shortcuts, which makes it difficult to match its constantly changing tangent directions as velocities evolve across timesteps, thereby leading to quality degradation. To address this limitation, we propose ArcFlow, a few-step distillation framework that explicitly employs non-linear flow trajectories to approximate pre-trained teacher trajectories. Concretely, ArcFlow parameterizes the velocity field underlying the inference trajectory as a mixture of continuous momentum processes. This enables ArcFlow to capture velocity evolution and extrapolate coherent velocities to form a continuous non-linear trajectory within each denoising step. Importantly, this parameterization admits an analytical integration of this non-linear trajectory, which circumvents numerical discretization errors and results in high-precision approximation of the teacher trajectory. To train this parameterization into a few-step generator, we implement ArcFlow via trajectory distillation on pre-trained teacher models using lightweight adapters. This strategy ensures fast, stable convergence while preserving generative diversity and quality. Built on large-scale models (Qwen-Image-20B and FLUX.1-dev), ArcFlow only fine-tunes on less than 5% of original parameters and achieves a 40x speedup with 2 NFEs over the original multi-step teachers without significant quality degradation. Experiments on benchmarks show the effectiveness of ArcFlow both qualitatively and quantitatively.

cs.CV↗