arXiv subjects

Haonan Wang

Publications and source records attributed to Haonan Wang.

At least 73 records · Page 4Linked to original sources

arXiv2025

Injecting Frame-Event Complementary Fusion into Diffusion for Optical Flow in Challenging Scenes

Optical flow estimation has achieved promising results in conventional scenes but faces challenges in high-speed and low-light scenes, which suffer from motion blur and insufficient illumination. These conditions lead to weakened texture and amplified noise and deteriorate the appearance saturation and boundary completeness of frame cameras, which are necessary for motion feature matching. In degraded scenes, the frame camera provides dense appearance saturation but sparse boundary completeness due to its long imaging time and low dynamic range. In contrast, the event camera offers sparse appearance saturation, while its short imaging time and high dynamic range gives rise to dense boundary completeness. Traditionally, existing methods utilize feature fusion or domain adaptation to introduce event to improve boundary completeness. However, the appearance features are still deteriorated, which severely affects the mostly adopted discriminative models that learn the mapping from visual features to motion fields and generative models that generate motion fields based on given visual features. So we introduce diffusion models that learn the mapping from noising flow to clear flow, which is not affected by the deteriorated visual features. Therefore, we propose a novel optical flow estimation framework Diff-ABFlow based on diffusion models with frame-event appearance-boundary fusion.

cs.CV↗

arXiv2025

Research on the Integration of Embodied Intelligence and Reinforcement Learning in Textual Domains

This article addresses embodied intelligence and reinforcement learning integration in the field of text processing, aiming to enhance text handling with more intelligence on the basis of embodied intelligence's perception and action superiority and reinforcement learning's decision optimization capability. Through detailed theoretical explanation and experimental exploration, a novel integration model is introduced. This model has been demonstrated to be very effective in a wide range oftext processing tasks, validating its applicative potential

cs.CL↗

arXiv2025

ByteSized32Refactored: Towards an Extensible Interactive Text Games Corpus for LLM World Modeling and Evaluation

Simulating interactive world models remains a core challenge in Large Language Models(LLMs). In this work, we introduce the ByteSized32Refactored, a refactored, modular, and extensible implementation of the original ByteSized32 corpus to explore the task of text game generation. We further optimize the code structure of each text game and create the GameBasic.py foundation library, which centralizes common logic across all 32 games by abstracting 7 base classes (GameObject, etc.) into reusable modules, thereby reducing from 20k to 10k total lines of Python code compared to the original Bytesized32. Our refactored implementation enables extendability - with our centralized design, ByteSized32Refactored can be more efficiently extended to include text games of new scenarios and specifications by reusing the shared logic and functionalities. Extensive experiments with GPT-4o demonstrate a mix of performance - with Bytesized32Refactored, the generated text games for unseen scenarios showcase quality improvements on two of the four evaluation dimensions while decreases on the other two, indicating that the hierarchical structure of the refactored code presents new challenges for LLMs. Overall, we highlight that our extensible code structure, centered on the foundation library and the modular optimization, not only facilitates LLM adaptation to environment specifications but also establishes a scalable environment that supports future extensions.

cs.CL↗

arXiv2025

From Harm to Help: Turning Reasoning In-Context Demos into Assets for Reasoning LMs

Recent reasoning LLMs (RLMs), especially those trained with verifier-based reinforcement learning, often perform worse with few-shot CoT than with direct answering. We revisit this paradox using high-quality reasoning traces from DeepSeek-R1 as demonstrations and find that adding more exemplars consistently degrades accuracy, even when demonstrations are optimal. A detailed analysis reveals two mechanisms behind this decline: (i) semantic misguidance, where high textual similarity leads the model to treat the target as the same as the exemplar and to copy intermediate steps verbatim; and (ii) strategy transfer failure, where the model struggles to extract useful reasoning strategies and apply them to target questions. Guided by these, we introduce Insight-to-Solve (I2S), a sequential test-time procedure that turns demonstrations into explicit, reusable insights and derives a target-specific reasoning trace; optionally, the reasoning is self-refined for coherence and correctness (I2S+). Extensive experiments on diverse benchmarks show that I2S and I2S+ consistently outperform both direct answering and test-time scaling baselines across open- and closed-source models. Even for GPT models, our method helps: on AIME'25, GPT-4.1 rises by +14.0%, and o1-mini improves by +2.7% on AIME and +1.7% on GPQA, indicating that in-context demonstrations can be harnessed effectively via insight-refine-solve framework.

cs.CL↗

arXiv2025

DD-Ranking: Rethinking the Evaluation of Dataset Distillation

In recent years, dataset distillation has provided a reliable solution for data compression, where models trained on the resulting smaller synthetic datasets achieve performance comparable to those trained on the original datasets. To further improve the performance of synthetic datasets, various training pipelines and optimization objectives have been proposed, greatly advancing the field of dataset distillation. Recent decoupled dataset distillation methods introduce soft labels and stronger data augmentation during the post-evaluation phase and scale dataset distillation up to larger datasets (e.g., ImageNet-1K). However, this raises a question: Is accuracy still a reliable metric to fairly evaluate dataset distillation methods? Our empirical findings suggest that the performance improvements of these methods often stem from additional techniques rather than the inherent quality of the images themselves, with even randomly sampled images achieving superior results. Such misaligned evaluation settings severely hinder the development of DD. Therefore, we propose DD-Ranking, a unified evaluation framework, along with new general evaluation metrics to uncover the true performance improvements achieved by different methods. By refocusing on the actual information enhancement of distilled datasets, DD-Ranking provides a more comprehensive and fair evaluation standard for future research advancements.

cs.CV↗

arXiv2025

Bichromatic Moiré Superlattices for Tunable Quadrupolar Trions and Correlated States

Moiré superlattices in transition metal dichalcogenide heterostructures provide a platform to engineer many-body interactions. Here, we realize a bichromatic moiré superlattice in an asymmetric WSe$_2$/WS$_2$/WSe$_2$ heterotrilayer by combining R- and H-stacked bilayers with mismatched moiré wavelengths. This structure hosts fermionic quadrupolar moiré trions -- interlayer excitons bound to an opposite-layer hole -- with vanishing dipole moments. These trions arise from hybridized moiré potentials enabling multiple excitonic orbitals with tunable interlayer coupling, allowing control of excitonic and electronic ground states. We show that an out-of-plane electric field could effectively reshape moiré excitons and interlayer-intralayer electron correlations, driving a transition from interlayer to intralayer Mott states with enhanced Coulomb repulsion. The asymmetric stacking further enriches excitonic selection rules, broadening opportunities for spin-photon engineering. Our results demonstrate bichromatic moiré superlattices as a reconfigurable platform for emergent quantum states, where quadrupolar moiré trion emission may enable coherent and entangled quantum light manipulation.

cond-mat.mes-hall↗

arXiv2025

Valley-Selective Linear Dichroism and Excitonic Effects in Lieb-Lattice Altermagnets

Altermagnets have recently been recognized as a distinct class of magnetic materials characterized by alternative spin-split electronic structures without net magnetization. Despite intensive studies on their single-particle spintronic and valleytronic properties, many-electron interactions and optical responses of altermagnets remain less explored. In this work, we employ many-body perturbation theory to investigate excited states and their strain tunability. Using monolayer Mn2WS4 as a representative candidate, we uncover a novel spin valley-dependent excitonic selection rule in two-dimensional altermagnetic Lieb lattices. In addition to strongly bound excitons, we find that linearly polarized light selectively excites valley spin-polarized excitons. Moreover, due to the interplay between altermagnetic spin symmetry and electronic orbital character, we predict that applying uniaxial strain can lift valley degeneracy and enable the selective excitation of spin-polarized excitons, an effect not achievable in previously studied transition-metal dichalcogenides. These spin-valley-locked excitonic states and their strain tunability offer a robust mechanism for four-fold symmetric altermagnets to encode, store, and read valley/spin information.

cond-mat.mtrl-sci↗

arXiv2025

Axial Hall Effect in Altermagnetic Lieb Lattices

We predict a so-called axial Hall effect, a Berry-curvature-driven anomalous Hall response, in Lieb-lattice altermagnets. By constructing a tight-binding model, we identify the axial direction as a hidden topological degree of freedom. Breaking the double degeneracy of axial symmetry generates substantial Berry curvature and induces a pronounced anomalous Hall conductivity. First-principles calculations further confirm the emergence of this effect in strained altermagnets, particularly in ternary transition-metal dichalcogenides. We take Mn2WS4 as an example to reveal that the axial Hall effect originates from the interplay between Dresselhaus spin-orbit coupling and the intrinsic piezomagnetic response of Lieb-lattice altermagnets, leading to highly localized and enhanced Berry curvature. Remarkably, the magnitude of the axial Hall effect is significant and remains unchanged when varying the strain, highlighting the topological nature of the axial degree of freedom. Finally, in multilayer systems, the effect manifests as a distinctive thickness-dependent modulation of both anomalous and spin Hall responses. These findings emphasize the critical role of spin-orbit coupling and noncollinear spin textures in altermagnets, an area that has received limited attention, and open new pathways for exploring intrinsic Hall phenomena in topological magnetic systems.

cond-mat.mes-hall↗

arXiv2025

Design and Optimization of Reinforcement Learning-Based Agents in Text-Based Games

As AI technology advances, research in playing text-based games with agents has becomeprogressively popular. In this paper, a novel approach to agent design and agent learning ispresented with the context of reinforcement learning. A model of deep learning is first applied toprocess game text and build a world model. Next, the agent is learned through a policy gradient-based deep reinforcement learning method to facilitate conversion from state value to optimal policy.The enhanced agent works better in several text-based game experiments and significantlysurpasses previous agents on game completion ratio and win rate. Our study introduces novelunderstanding and empirical ground for using reinforcement learning for text games and sets thestage for developing and optimizing reinforcement learning agents for more general domains andproblems.

cs.CL↗

arXiv2025

Compressed Zeroth-Order Algorithm for Stochastic Distributed Nonconvex Optimization

This paper studies the stochastic distributed nonconvex optimization problem over a network of agents, where agents only access stochastic zeroth-order information about their local cost functions and collaboratively optimize the global objective over bandwidth-limited communication networks. To mitigate communication overhead and handle the unavailability of explicit gradient information, we propose a communication compressed zeroth-order stochastic distributed (CZSD) algorithm. By integrating a generalized contractive compressor and a stochastic two-point zeroth-order oracle, CZSD achieves convergence rates comparable to its exact communication counterpart while reducing both communication overhead and sampling complexity. Specifically, to the best of our knowledge, CZSD is the first compressed zeroth-order algorithm achieving linear speedup, with convergence rates of $\mathcal{O}(\sqrt{p}/\sqrt{nT})$ and $\mathcal{O}(p/(nT))$ under general nonconvex settings and the Polyak--Łojasiewicz condition, respectively. Numerical experiments validate the algorithm's effectiveness and communication efficiency.

math.OC↗

arXiv2025

Sliding two-dimensional superconductivity and charge-density-wave state in a bulk crystal

Superconductivity in the two-dimensional (2D) limit is a fertile ground for exotic quantum phenomena-many of which remain elusive in their 3D counterparts. While studies of 2D superconductivity have predominantly focused on mono- or few-layer systems, we demonstrate an alternative route-interlayer sliding in bulk crystals. Through a precisely controlled growth strategy, we engineer interlayer sliding in bulk 3R-NbSe2, deliberately disrupting [001] mirror symmetry and drastically suppressing interlayer coupling. Remarkably, this structural manipulation stabilizes Ising-type superconductivity coexisting with an unconventional charge-density-wave (CDW) state akin to that of monolayer 2H-NbSe2. The sliding phase exhibits a pronounced suppression of the upper critical field at low temperatures, revealing a delicate competition between Ising and Rashba spin-orbit coupling (SOC) in the globally noncentrosymmetric lattice. Intriguingly, the superconducting state displays two-fold symmetry, a signature that may arise from asymmetric SOC or a multi-component pairing order parameter. Our work establishes interlayer sliding as a symmetry-breaking tool to promote 2D superconductivity in bulk materials-without resorting to extrinsic intercalation or doping. More broadly, this approach sets a paradigm for unlocking hidden quantum states in layered materials, offering a new dimension in design of quantum matter.

cond-mat.supr-con↗

arXiv2025

LoReUn: Data Itself Implicitly Provides Cues to Improve Machine Unlearning

Recent generative models face significant risks of producing harmful content, which has underscored the importance of machine unlearning (MU) as a critical technique for eliminating the influence of undesired data. However, existing MU methods typically assign the same weight to all data to be forgotten, which makes it difficult to effectively forget certain data that is harder to unlearn than others. In this paper, we empirically demonstrate that the loss of data itself can implicitly reflect its varying difficulty. Building on this insight, we introduce Loss-based Reweighting Unlearning (LoReUn), a simple yet effective plug-and-play strategy that dynamically reweights data during the unlearning process with minimal additional computational overhead. Our approach significantly reduces the gap between existing MU methods and exact unlearning in both image classification and generation tasks, effectively enhancing the prevention of harmful content generation in text-to-image diffusion models.

cs.LG↗

arXiv2025

TrackAny3D: Transferring Pretrained 3D Models for Category-unified 3D Point Cloud Tracking

3D LiDAR-based single object tracking (SOT) relies on sparse and irregular point clouds, posing challenges from geometric variations in scale, motion patterns, and structural complexity across object categories. Current category-specific approaches achieve good accuracy but are impractical for real-world use, requiring separate models for each category and showing limited generalization. To tackle these issues, we propose TrackAny3D, the first framework to transfer large-scale pretrained 3D models for category-agnostic 3D SOT. We first integrate parameter-efficient adapters to bridge the gap between pretraining and tracking tasks while preserving geometric priors. Then, we introduce a Mixture-of-Geometry-Experts (MoGE) architecture that adaptively activates specialized subnetworks based on distinct geometric characteristics. Additionally, we design a temporal context optimization strategy that incorporates learnable temporal tokens and a dynamic mask weighting module to propagate historical information and mitigate temporal drift. Experiments on three commonly-used benchmarks show that TrackAny3D establishes new state-of-the-art performance on category-agnostic 3D SOT, demonstrating strong generalization and competitiveness. We hope this work will enlighten the community on the importance of unified models and further expand the use of large-scale pretrained models in this field.

cs.CV↗

arXiv2025

Giant plateau-like topological Hall effect controlled by tailoring the magnetic exchange stiffness in a kagome magnet

The ferrimagnet TbMn6Sn6 has attracted vast attention, because its pristine Mn kagome lattice with strong spin-orbit coupling and out-of-plane Tb-Mn exchange supports quantum-limit Chern topological magnetism which can be described by the simple spinless Haldane model. We unveil herein that engineering the kagome lattice through partial substitution of Mn with nonmagnetic Cr induces a striking structural reorganization-Cr preferentially concentrates within a single Mn layer per unit cell, reducing the crystal symmetry from the D6h point group to the C2. This tailored structure configuration gives rise to a plateau-like topological Hall effect (THE), achieving a record-breaking resistivity of 19.1 ohm cm among bulk systems. Complementary magnetic force microscopy measurements unveil a magnetic domain transition near 1 T at 180 K, aligning with the field-dependent phase diagram of the THE. Our direct visualization of the magnetic domain structure underscores the critical role of broken kagome lattice symmetry in generating distinct exchange stiffness between the two Mn layers. These findings establish a new paradigm for exploring exotic states in kagome topological magnets and provide a proof-of-principle strategy for unraveling the interplay between magnetism and emergent topological properties in kagome systems.

cond-mat.str-el↗

arXiv2025

PromptArmor: Simple yet Effective Prompt Injection Defenses

Despite their potential, recent research has demonstrated that LLM agents are vulnerable to prompt injection attacks, where malicious prompts are injected into the agent's input, causing it to perform an attacker-specified task rather than the intended task provided by the user. In this paper, we present PromptArmor, a simple yet effective defense against prompt injection attacks. Specifically, PromptArmor prompts an off-the-shelf LLM to detect and remove potential injected prompts from the input before the agent processes it. Our results show that PromptArmor can accurately identify and remove injected prompts. For example, using GPT-4o, GPT-4.1, or o4-mini, PromptArmor achieves both a false positive rate and a false negative rate below 1% on the AgentDojo benchmark. Moreover, after removing injected prompts with PromptArmor, the attack success rate drops to below 1%. We also demonstrate PromptArmor's effectiveness against adaptive attacks and explore different strategies for prompting an LLM. We recommend that PromptArmor be adopted as a standard baseline for evaluating new defenses against prompt injection attacks.

cs.CR↗

arXiv2025

Neurons: Emulating the Human Visual Cortex Improves Fidelity and Interpretability in fMRI-to-Video Reconstruction

Decoding visual stimuli from neural activity is essential for understanding the human brain. While fMRI methods have successfully reconstructed static images, fMRI-to-video reconstruction faces challenges due to the need for capturing spatiotemporal dynamics like motion and scene transitions. Recent approaches have improved semantic and perceptual alignment but struggle to integrate coarse fMRI data with detailed visual features. Inspired by the hierarchical organization of the visual system, we propose NEURONS, a novel framework that decouples learning into four correlated sub-tasks: key object segmentation, concept recognition, scene description, and blurry video reconstruction. This approach simulates the visual cortex's functional specialization, allowing the model to capture diverse video content. In the inference stage, NEURONS generates robust conditioning signals for a pre-trained text-to-video diffusion model to reconstruct the videos. Extensive experiments demonstrate that NEURONS outperforms state-of-the-art baselines, achieving solid improvements in video consistency (26.6%) and semantic-level accuracy (19.1%). Notably, NEURONS shows a strong functional correlation with the visual cortex, highlighting its potential for brain-computer interfaces and clinical applications. Code and model weights are available at: https://github.com/xmed-lab/NEURONS.

cs.CV↗

arXiv2025

STD-GS: Exploring Frame-Event Interaction for SpatioTemporal-Disentangled Gaussian Splatting to Reconstruct High-Dynamic Scene

High-dynamic scene reconstruction aims to represent static background with rigid spatial features and dynamic objects with deformed continuous spatiotemporal features. Typically, existing methods adopt unified representation model (e.g., Gaussian) to directly match the spatiotemporal features of dynamic scene from frame camera. However, this unified paradigm fails in the potential discontinuous temporal features of objects due to frame imaging and the heterogeneous spatial features between background and objects. To address this issue, we disentangle the spatiotemporal features into various latent representations to alleviate the spatiotemporal mismatching between background and objects. In this work, we introduce event camera to compensate for frame camera, and propose a spatiotemporal-disentangled Gaussian splatting framework for high-dynamic scene reconstruction. As for dynamic scene, we figure out that background and objects have appearance discrepancy in frame-based spatial features and motion discrepancy in event-based temporal features, which motivates us to distinguish the spatiotemporal features between background and objects via clustering. As for dynamic object, we discover that Gaussian representations and event data share the consistent spatiotemporal characteristic, which could serve as a prior to guide the spatiotemporal disentanglement of object Gaussians. Within Gaussian splatting framework, the cumulative scene-object disentanglement can improve the spatiotemporal discrimination between background and objects to render the time-continuous dynamic scene. Extensive experiments have been performed to verify the superiority of the proposed method.

cs.CV↗

arXiv2025

Token Activation Map to Visually Explain Multimodal LLMs

Multimodal large language models (MLLMs) are broadly empowering various fields. Despite their advancements, the explainability of MLLMs remains less explored, hindering deeper understanding, model credibility, and effective visualization. Unlike conventional vision models (e.g., CNNs, ViTs, CLIP) that produce a single output, MLLMs generate sequences of tokens progressively, where each generated token depends on the previous context. Therefore, earlier context tokens can introduce redundant activations that interfere with the explanation of later tokens beyond their original information. Existing studies often overlook this issue, but our observations reveal that these redundant correlations can significantly hurt the reliability of explanations. To address this, we propose an estimated causal inference method to mitigate the interference of context to achieve high-quality MLLM explanation, with a novel rank Gaussian filter to further reduce activation noises. We term this method Token Activation Map (TAM) to highlight the consideration of interactions between tokens. TAM also indicates that it excels at explaining multiple tokens of MLLM, which is different from the Class Activation Map (CAM) for a single prediction. Our TAM method significantly outperforms existing SoTA methods, showcasing high-quality visualization results that can be utilized for various scenarios, such as object localization, failure case analysis, video visualization, MLLMs visual comparison, and model understanding (e.g., color, shape, action, location, visual reasoning, multi-turn conversation, etc). The code is available atgithub.com/xmed-lab/TAM.

cs.CV↗