Searcharxiv⌕ Search

arXiv subjects

Hongtao Wang

Publications and source records attributed to Hongtao Wang.

At least 19 recordsLinked to original sources

Optical Representation Neural Networks Enabled by Weakly Supervised Learning

Diffractive optical neural networks (ONNs) show remarkable promise in high-speed and energy-efficient optical computing. However, most existing ONNs are optimized to reproduce predefined output intensity distributions, with the desired optical responses explicitly specified before training. Here, we introduce optical representation neural networks (ORNNs) enabled by coarse-grained weak supervision. Using handwritten Devanagari, Arabic, and English characters as three semantic groups, a modified contrastive objective encourages samples from the same script to cluster into similar optical representations. Remarkably, ORNNs improve the classification accuracy by approximately 9% over conventional supervised ONNs, while requiring only around 18% of the training steps to reach a stable test accuracy of 84.5%. These features can be selectively interpreted using corresponding semantic decoders. Building on this concept, we further realized semantic optical encryption with electronic neural network (ENN) decryption using experimentally acquired representations from a single-layer diffractive network fabricated by two-photon polymerization lithography (TPL). Here, we combine information from different scripts into a mixed optical ciphertext that is then first optically classified before being decrypted with a matched ENN. This work broadens the training paradigm of ONNs by coupling upstream optical representation learning with downstream task-specific decoding, opening new avenues for optical information processing.

physics.optics↗

GAP: Generalized Adaptation Prior for Dexterous Piano Playing

Dexterous hands enable versatile manipulation, but policies trained in simulation often degrade on hardware owing to non-idealities, and the high-dimensional action space makes real-world adaptation costly. Piano playing, which requires precise and coordinated multifinger motion, offers a representative testbed. We propose GAP (Generalized Adaptation Prior for Dexterous Piano Playing), a sample-efficient adaptation framework that learns how to adapt, rather than how to act, during simulation pretraining and transfers this adaptation prior to hardware. Between repeated executions, GAP uses feedback to adjust the actions of the next one: a gradient guidance network learns from paired executions how adjusting actions affects playing performance, and its gradients guide these adjustments. In simulation, GAP yields approximately 2.4 times the F1 scores improvement of the existing fast-adaptation baseline for dexterous piano playing under the same interaction budget. The learned prior also generalizes to musical pieces held out from pretraining. On a physical dexterous hand, the same simulation-pretrained prior improves F1 scores by 3.6%-23% across three pieces of varying difficulty after approximately 7 min of adaptation each. These results suggest a practical route to reduce the real-world interaction needed to deploy dexterous hands.

cs.RO↗

Symmetry-governed nonlinear metasurface simultaneously enhances circular dichroism, helicity purity, and Q-factor

Circularly polarized light underpins a broad range of chiroptical and spin-selective photonic technologies, yet extending resonant chirality to nonlinear harmonic emission remains challenging. Here, we experimentally demonstrate a symmetry-governed mechanism that maps the maximal linear chirality of high-Q degenerate resonances onto the polarization of the generated harmonic. Preserved rotational symmetry protects modal degeneracy and polarization-selection rules, while broken mirror symmetries enable a strong chiroptical response. The metasurface thereby filters the pump polarization and purifies the handedness of the nonlinear output, simultaneously achieving third-harmonic circular dichroism of -0.98, a degree of circular polarization of 0.83, and an effective nonlinear Q-factor of 140. Importantly, the nonlinear emission preserves its handedness even under spatially depolarized excitation, demonstrating robustness to variations in the incident polarization state. These results establish complementary symmetry preservation and breaking as a deterministic route for transferring resonant chirality from linear to nonlinear optical responses, providing a general design principle for nonlinear chiral photonics.

physics.optics↗

SLDR: Defending Against Malicious Fine-tuning via Selective Layers Recovery and Dynamic Routing

Fine-tuning-as-a-service enables users to adapt aligned large language models (LLMs) to specialized tasks, but malicious fine-tuning can erode refusal behavior while preserving task performance on legitimate inputs. We revisit recent layer-wise safety diagnostics and find that safety sensitivity is signed: scaling different layers can strengthen refusal, weaken it, or have little effect. Motivated by this observation, we propose SLDR, a post-fine-tuning defense based on Selective Layers Recovery and Dynamic Routing. SLDR trains a LoRA recovery adapter only on the layers with the maximum and minimum sensitivity scores in the signed spectrum, and uses representation-based dynamic routing inference to activate the adapter only for malicious queries. Across four model architectures, five downstream tasks, and four harmful benchmarks, SLDR substantially reduces harmful outputs while preserving downstream utility. On Llama3.1/SST2, SLDR reduces the average harmful score from 11.54 to 0.08 while maintaining downstream accuracy, and the harmful score remains near zero under poisoning ratios up to 0.9. The code is available at https://github.com/Stardust457/SLDR.

cs.CR↗

MemPoison: Bypassing Selective Memory Mechanisms to Plant Backdoors in LLM Agents

Large language model (LLM) agents increasingly leverage long-term memory to support persistent and autonomous task execution. However, this capability also introduces a new attack surface: memory poisoning, where adversaries can inject malicious information to influence future behavior. Existing memory poisoning attacks often assume that injected content can be stored directly in memory, overlooking the selective extraction and rewriting stages in modern memory pipelines. This makes prior methods ineffective under realistic settings. In this paper, we propose MemPoison, a novel memory poisoning attack that bypasses selective memory mechanisms in LLM agents, where an attacker can inject triggerable backdoors into the agent's long-term memory through dialogue interactions, thereby misleading its subsequent responses. MemPoison introduces three key components: (i) a semantic relational bridge that binds the trigger and payload into a coherent statement to ensure their co-extraction into memory; (ii) entity masquerading that optimizes triggers to mimic named entities, resisting rewriting; and (iii) joint embedding optimization that shapes trigger-injected texts into a tight cluster in the embedding space while maintaining isolation from benign embeddings for stealth. Evaluations across different agent domains and memory mechanisms show MemPoison achieves attack success rates up to 0.95, outperforming existing baselines. Mechanistic analysis reveals the exploitation of embedding-space anisotropy and attention redistribution, highlighting core vulnerabilities in selective memory systems. We evaluate multiple defense strategies and demonstrate their fundamental limitations in mitigating the attack. We release our code to facilitate further research.

cs.CR↗

TAGA: Terrain-aware Active Gaze Learning for Generalizable Agile Humanoid Locomotion

Agile humanoid locomotion across diverse challenging terrain demands both wide perceptual coverage and precise local geometry understanding. Motivated by the way humans selectively look at relevant terrain during locomotion, we introduce TAGA, a Terrain-aware Active Gaze learning framework for Attention-based humanoid control. By fusing vision, proprioception, and motion commands, our framework guides the model to learn anticipatory cues and actively attend to specific areas of the height scan, selectively using these informative regions for the downstream network. This adaptively increases the information density of observations under tight onboard computational constraints, thus enabling fine-grained perceptive locomotion over larger-scale terrains. We find that such gaze behaviors can naturally emerge through reinforcement learning alone, without requiring additional supervision or explicit guidance, significantly improving training efficiency. As a result, the trained policy demonstrates robust and generalizable locomotion in simulation and on hardware, including reliable terrain-aware foothold selection, elevated-platform traversal, competitive sparse-foothold traversal, and the largest reported real-world gap traversal distance of 1.2m among perceptive humanoid locomotion systems, while maintaining stability under severe perceptual disturbances and environmental interference.

cs.RO↗

PGMT: Perceptive General Motion Tracking for Humanoid Robots

Humanoid motion trackers can reproduce diverse whole-body motions, but their performance degrades on complex terrain where terrain-agnostic references become physically infeasible. We present PGMT, a Perceptive General Motion Tracking pipeline for humanoid robots that learns terrain adaptation from independently selected motion references and terrains. PGMT first learns a general tracking and recovery prior, then incorporates terrain perception through motion-conditioned terrain glimpses that selectively encode regions relevant to the current motion. Terrain-aware tracking relaxation allows necessary deviations from the reference while preserving its motion intent. Zero-shot deployment on a Unitree G1 demonstrates robust terrain-adaptive locomotion and whole-body motion execution over real-world terrain with obstacles up to 37 cm high, while supporting teleoperation, dynamic motion tracking, and fall recovery. PGMT extends general humanoid motion tracking beyond flat ground, providing a unified policy for terrain-adaptive locomotion, diverse whole-body behaviors, and teleoperation in complex environments. Project homepage: https://luyili.github.io/pgmt/

cs.RO↗

Expressive Robotic Pianist: Mastering Complex Piano Repertoire with Graph-Mimic and Musical Dynamics

Enabling robots to perform musical instruments with human-level expressivity represents a frontier in bridging the gap between mechanical precision and artistic interpretation. Despite advances in robotic dexterity, replicating the fluid finger transitions and nuanced dynamic control characteristic of human pianists remains a significant challenge. Through a reinforcement learning-based control framework, we demonstrate that a dexterous robotic hand can achieve high-fidelity performance across a diverse piano repertoire. Central to our approach is a graph-based optimization strategy that guides the robot to generate natural pre-press and key-press fingering strategies that closely resemble human movement patterns. To achieve expressive sound production, the control system is coupled with a physics-inspired acoustic model that modulates keypress velocity to accurately reproduce the dynamic variations specified in musical scores. Quantitative evaluations demonstrate that our expressive control model significantly outperforms baseline methods in both finger morphology similarity and dynamic velocity accuracy. In a perceptual test involving participants from diverse listener groups, performances generated by our system are significantly preferred over baseline robotic performances and are indistinguishable from human performances for non-professional audiences. Furthermore, extensive experiments across multiple musical styles confirm that our method maintains high note-level accuracy while achieving expressive performance. Our approach provides a robust pathway for robotic systems to move beyond mere mechanical accuracy, elevating robotic musicianship to a level of expressive performance comparable to human pianists.

cs.RO↗

Breaking speed scaling in quadrupedal robots via Huygens' coupled-pendulum dynamics

Achieving biological-level running speeds has largely been pursued through advances in control algorithms, which improve the utilization of existing hardware. However, the ultimate speed limits remain governed by the underlying force and torque requirements of rapid locomotion, which are typically addressed through increased actuator capacity. Inspired by Huygens' coupled pendulums, we demonstrate that superior locomotion can emerge from principled exploitation of intrinsic dynamics rather than brute-force hardware scaling. Inter-limb inertial coupling redistributes energy across the gait cycle and reduces peak joint torque required for rapid periodic motion, thereby expanding the achievable speed without proportional increases in actuator capability. Incorporating hardware parameters as additional design variables further extends this analysis into a co-optimization framework, enabling the systematic utilization of inertial coupling in robot design. Guided by this framework, a quadruped robot achieves a running speed of 10.74 m/s (Froude number 21.4) and completes a 100-meter sprint in 12.2 seconds, representing the first legged robot to surpass 10 m/s. These results establish inertial coupling as an underlying mechanism governing high-speed legged locomotion and highlight its role in reducing force requirements, offering new insights into the design of agile robotic systems.

cs.RO↗

Extending the Speed Limit of Quadrupedal Locomotion via Refined Actuator Modeling and Adaptive Command Scheduling

Achieving high-speed locomotion in quadrupedal robots remains highly challenging, as actuators operate near their physical limits and exhibit pronounced nonlinearities. However, many existing methods neglect actuator nonlinearities and physical constraints during training, leading to a significant sim-to-real gap under highly dynamic motions and limiting achievable performance. To address this issue, we propose a high-speed locomotion framework that reduces sim-to-real discrepancies and stabilizes learning over a wide command distribution. A refined actuator model explicitly captures high-speed voltage coupling and magnetic saturation, enabling a more accurate representation of the torque-speed envelope. In addition, a reinforcement learning framework incorporating a two-stage curriculum and adaptive command scheduling (ACS) ensures stable training. Experiments on the 36.5 kg quadruped BlackPanther2 (BP2) demonstrate speeds of up to 13.2 m/s on a treadmill and 11.65 m/s outdoors, establishing a new state-of-the-art and, to the best of our knowledge, a world record for quadrupedal robot locomotion. The results further highlight the importance of accurate actuator modeling in preventing non-physical policy exploitation, and show that ACS improves robustness without sacrificing performance.

cs.RO↗

Sublime Transfer Printing of Three-Dimensional Nanostructure Ensembles

High-resolution three-dimensional (3D) nanostructures for visible-light photon manipulation provide unique and bespoke capabilities in optics and photonics. However subwavelength nanofabrication and reliable ensemble manipulation of the 3D prints onto arbitrary substrates remain challenging. Here, we introduce sublime transfer strategy tailored for transfer printing ensembles of delicate 3D printed nanostructures. This strategy enables conformal, damage-free integration of arrays of 3D structures on diverse substrates. Naphthalene acts as a transient stamp to encapsulate the structures during transfer and placement. We rely on the low sublimation temperature of naphthalene to release the structures reliably with nearly zero stress, preventing mechanical damage and positional misalignment. This approach is broadly applicable to integrate diverse nanostructures and photonic devices onto various substrates, and enabling inorganic architectures through ensemble uniform post-processing, including 2.5D photonic crystals on flexible PDMS, diffractive optical elements on curved lenses, spiral phase plates on CMOS chips, multilayer achromatic metalens on optical fiber facet, as well as 3D glass photonic crystals and optical topological resonators on anti-stiction quartz.

physics.optics↗

SWAP: Symmetric Equivariant World-Model for Agile Robot Parkour

While latent world models enable the proactive predictions required for extreme parkour, their purely data-driven nature forces them to redundantly encode left-right symmetric interactions as independent patterns. This inflates the learning burden and hinders the capture of geometric regularities, restricting the latent space's efficiency for downstream policies. To address this, we propose SWAP, an end-to-end equivariant symmetric world model. This framework embeds symmetry directly into both the world model and the actor-critic networks. In real-world tests, the robot leaps across a 2.13 m gap and climbs a 1.63 m platform, breaking records for quadruped parkour. Furthermore, the framework exhibits robust geometric generalization to unseen mirrored terrains and exceptional zero-shot transferability across diverse outdoor environments. These results demonstrate that symmetry equivariance is an effective structural prior for pushing the physical boundaries of learned legged locomotion.

cs.RO↗

NILC: Discovering New Intents with LLM-assisted Clustering

New intent discovery (NID) seeks to recognize both new and known intents from unlabeled user utterances, which finds prevalent use in practical dialogue systems. Existing works towards NID mainly adopt a cascaded architecture, wherein the first stage focuses on encoding the utterances into informative text embeddings beforehand, while the latter is to group similar embeddings into clusters (i.e., intents), typically by K-Means. However, such a cascaded pipeline fails to leverage the feedback from both steps for mutual refinement, and, meanwhile, the embedding-only clustering overlooks nuanced textual semantics, leading to suboptimal performance. To bridge this gap, this paper proposes NILC, a novel clustering framework specially catered for effective NID. Particularly, NILC follows an iterative workflow, in which clustering assignments are judiciously updated by carefully refining cluster centroids and text embeddings of uncertain utterances with the aid of large language models (LLMs). Specifically, NILC first taps into LLMs to create additional semantic centroids for clusters, thereby enriching the contextual semantics of the Euclidean centroids of embeddings. Moreover, LLMs are then harnessed to augment hard samples (ambiguous or terse utterances) identified from clusters via rewriting for subsequent cluster correction. Further, we inject supervision signals through non-trivial techniques seeding and soft must links for more accurate NID in the semi-supervised setting. Extensive experiments comparing NILC against multiple recent baselines under both unsupervised and semi-supervised settings showcase that NILC can achieve significant performance improvements over six benchmark datasets of diverse domains consistently.

cs.CL↗

Adaptive Graph Refinement and Label Propagation with LLMs for Cost-Effective Entity Resolution

Dirty entity resolution (ER), which identifies records referring to the same real-world entity from a single, messy dataset, is a fundamental task in data management and mining. However, the dominant blocking-matching-clustering paradigm for ER suffers from critical flaws. Its cascaded, decoupled workflow essentially produces a static, sparse graph plagued by missing edges (due to blocking failures) and noisy links (due to matching errors), causing error propagation and yielding suboptimal clusters, particularly when rigid transitivity is imposed in the clustering. We contend that matching and clustering are fundamentally synergistic, both optimizing for the construction of an ideal entity graph. Building upon this insight, we propose Alper, a unified framework that integrates these steps into an iterative probabilistic label propagation process over a global, evolving graph. Unlike disjoint blocking, Alper refines the graph structure and labels dynamically by adaptively integrating "weak but cheap" signals from graph propagation with "strong but expensive" LLM-based pairwise queries. For higher cost-effectiveness, we formulate the signal selection as a constrained optimization problem maximizing cumulative marginal gain under a query budget, solved via our greedy algorithm with provable theoretical guarantees. Our extensive experiments over eight benchmark datasets demonstrate that Alper is consistently superior to state-of-the-art cascaded pipelines.

cs.CL↗

DuoTeach: Dual Role Self-Teaching for Coarse-to-Fine Decision Coordination in Vision--Language Models

Coarse-to-fine path decision-making requires predicting a valid taxonomy path in which earlier decisions constrain later ones. However, existing benchmarks score each level independently, obscuring cross-level validity and consistency. To better align evaluation with this setting, we introduce a Joint Path Decision (JPD) protocol that requires predicting the full path in one call, together with Depth-Weighted Prefix Accuracy (DWPA), a metric family that measures path reliability with tunable emphasis on deeper levels. Under JPD, strong vision-language models (VLMs) frequently produce invalid parent-child pairs and brittle full-path predictions, suggesting that their failures stem not only from incomplete taxonomic knowledge but also from unstable cross-level decision coordination. To address this problem, we propose DuoTeach, a dual-role self-teaching distillation framework that requires no ground-truth labels and reuses the same pretrained VLM in two roles. Its Decision-Conditioned Rollout (DCR) generates more coherent teacher traces by conditioning each level on prior decisions, and distills this coordinated behavior into the student without additional test-time rollouts. Across multiple taxonomy-structured benchmarks and VLM base models, DuoTeach improves in-domain DWPA (alpha = 0.95) by up to 30.24 points and boosts zero-shot performance on unseen taxonomies from 17.17% to 43.66%. Further analyses attribute these gains to improved within-call multi-level decision coordination.

cs.MM↗

Identifying Compton-thick active galactic nuclei in the COSMOS. II. Searching among mid-infrared selected AGNs

Compton-thick active galactic nuclei (CT-AGNs), defined by column density $\mathrm{N_H} \geqslant 1.5 \times 10^{24} \ \mathrm{cm}^{-2}$, are so heavily absorbed that their X-ray emission is often feeble, even undetectable by X-ray instruments. Nevertheless, their radiation is expected to be a substantial contributor to the cosmic X-ray background (CXB), predicting that CT-AGNs should comprise at least $\sim$30% of the total AGN population. In the Cosmological Evolution Survey (COSMOS), the identified CT-AGN fraction falls far below theoretical expectations, indicating that a substantial population of CT-AGNs is hidden due to their low photon counts or their flux below the current flux limits of X-ray instruments. This work focuses on identifying CT-AGNs hidden in mid-infrared (MIR)-selected AGNs. First, we selected a sample of 1,104 MIR-selected AGNs that were covered but individually undetected by X-ray. Next, we reduced the X-ray data in the COSMOS and analyzed multiwavelength data in our sample to derive the key physical parameters required for CT-AGN identification. Using MIR diagnostics, we first find out 7 to 23 CT-AGN candidates. Their subsequent X-ray stacking analysis reveals a clear detection at $>3σ$ significance in the soft band and only $>1σ$ significance in the hard band. We fit the stacked soft- and hard-band fluxes with a physical model and confirm that these sources are absorbed by Compton-thick material. However, CT-AGNs constitute only 2.1% (23/1104) of our sample, significantly below the fraction predicted by CXB synthesis models, indicating that a considerable population of CT-AGNs remains missed by our selection. A comparison of host-galaxy properties between CT-AGNs and non-CT-AGNs reveals no significant differences.

astro-ph.GA↗

Enhancing Volumetric Optical Chirality through 2D-3D Structural Design Evolution

Circular dichroism (CD) sensing plays a pivotal role in probing molecular chirality in biomedical sciences. However, engineering superchiral electromagnetic fields that can reliably amplify the faint signatures of chiral analytes remains profoundly challenging. Central to this difficulty is the need to balance two competing demands: maximizing the enhancement of chiral fields while maintaining a sufficiently large interaction volume for effective molecular interrogation. Here, we introduce a figure of merit (FOM) that captures the enhancement and spatial coverage of superchiral fields to benchmark different chiral-field configurations. We examine the effects of helix-geometry evolution on the FOM, including 2D to 3D chirality induction, winding-number escalation, helical-order enhancement, and transverse dilation. By tuning these structural degrees of freedom, the sensing volume can be enlarged without compromising the distribution and enhancement strength of fields. The optimized triple-strand helix markedly enhanced the analyte CD signal, yielding a FOM of 2.43*10^10 nm3, which surpassed prior 2D and 3D configurations by over an order of magnitude. The proposed FOM exhibits a strong linear correlation (R^2 = 0.9256) with the analyte CD signal. Our findings provide a systematic design framework for 3D chiral structures and a robust metric for assessing their chiroptical sensing performance, particularly in scenarios involving clusters of randomly oriented small molecules or a large chiral molecule.

physics.optics↗

Cross-Contrastive Clustering for Multimodal Attributed Graphs with Dual Graph Filtering

Multimodal Attributed Graphs (MMAGs) are an expressive data model for representing the complex interconnections among entities that associate attributes from multiple data modalities (text, images, etc.). Clustering over such data finds numerous practical applications in real scenarios, including social community detection, medical data analytics, etc. However, as revealed by our empirical studies, existing multi-view clustering solutions largely rely on the high correlation between attributes across various views and overlook the unique characteristics (e.g., low modality-wise correlation and intense feature-wise noise) of multimodal attributes output by large pre-trained language and vision models in MMAGs, leading to suboptimal clustering performance. Inspired by foregoing empirical observations and our theoretical analyses with graph signal processing, we propose the Dual Graph Filtering (DGF) scheme, which innovatively incorporates a feature-wise denoising component into node representation learning, thereby effectively overcoming the limitations of traditional graph filters adopted in the extant multi-view graph clustering approaches. On top of that, DGF includes a tri-cross contrastive training strategy that employs instance-level contrastive learning across modalities, neighborhoods, and communities for learning robust and discriminative node representations. Our comprehensive experiments on eight benchmark MMAG datasets exhibit that DGF is able to outperform a wide range of state-of-the-art baselines consistently and significantly in terms of clustering quality measured against ground-truth labels.

cs.LG↗