Searcharxiv⌕ Search

arXiv subjects

Miao Pan

Publications and source records attributed to Miao Pan.

At least 19 recordsLinked to original sources
arXiv2026

A Novel Semantic Manifold Alignment Attack against Embedding-to-Embedding Obfuscation in Privacy-Preserving LLMs

With the widespread applications of large language models (LLMs), privacy-preserving inference has become increasingly essential for sensitive queries. To balance privacy and utility, a series of lightweight obfuscation approaches has recently been proposed, where users locally transform plaintext embeddings into the fixed ciphertext ones. While such Embedding-to-Embedding Obfuscation (E2EO) schemes demonstrate considerable resilience against traditional token frequency and embedding inversion attacks, the core mechanism behind remains to be the large-scale one-to-one substitution, which provides no cryptographic guarantees. In this paper, we propose Proxy Manifold Alignment (PMA), a novel attack against E2EO in privacy-preserving LLMs. Our key observation is that E2EO schemes keep the original semantic structure, so that the obfuscated vector stream can be regarded as an unknown tokenizer-language whose symbols are the vectors themselves. Therefore, the proposed ciphertext to plaintext reconstruction attack can be formulated as a translation task from the unknown tokenizer-language to plaintext. Specifically, by only accessing the obfuscated vector stream, the target tokenizer and a public corpus, the PMA attack first employs Word2Vec to model the co-occurrence patterns within the obfuscated stream and the public corpus independently, and constructs two proxy vector embeddings. Then, the attack aligns the underlying manifolds of these two embeddings based on structural similarity. Finally, it maps the obfuscated vectors back to plaintext. Experimental results demonstrate that PMA consistently achieves higher plaintext recovery than other state-of-the-art attack methods.

cs.CR↗
arXiv2026

PhysWave: Physics-Guided Latent Diffusion Models for Controllable Spatial Audio Generation

Text-to-spatial audio generation, such as text-to-First-Order Ambisonics (FOA), provides a convenient way to create spatial audio for billion-dollar gaming and film industries. However, existing text-to-FOA methods are largely data-driven and may produce audio that violates acoustic relations between source direction and distance. They also separate descriptive and parametric control, forcing users to trade usability for precision. In this paper, we present PhysWave, a physics-guided latent diffusion model for controllable text-to-FOA generation. PhysWave unifies natural-language and trajectory control through a shared waypoint-caption representation, and augments diffusion training with two differentiable acoustic priors: spherical-harmonic direction consistency and inverse-square distance consistency. To support dynamic spatial generation, we further construct a 300K-clip FOA dataset with diverse sound categories and source trajectories. Extensive results show that the proposed priors help PhysWave generate spatially consistent FOA audio while maintaining competitive audio quality. Further analyses show that these physics priors improve spatial consistency during training and can also be used as inference-time guidance for training-free spatial refinement.

cs.SD↗
arXiv2026

Embodied-Navigator: Point, Think, Memorize, and Align for Efficient Navigation

Although Large Vision-Language Models (VLMs) have significantly advanced embodied navigation, their direct deployment remains challenging, as existing methods often force VLMs into unnatural action spaces that misalign with their 2D pre-training priors, compounded by rigid reasoning schedules and inefficient memory management. To overcome these limitations, we propose TAMP-Nav, a unified framework for efficient embodied navigation. First, we introduce a Pixel-to-3D Action Formulation (Point) that reformulates navigation into 2D visual prompting. Specifically, the VLM merely selects 2D pixels, which are then projected into 3D coordinates for a low-level SLAM controller. This design naturally aligns embodied execution with the VLM's inherent 2D visual capabilities. Second, we propose an integrated Selective Reasoning and Anchor-Trajectory Memory mechanism (Think and Memorize), which dynamically triggers Chain-of-Thought and retains high-fidelity memory only at critical nodes, compressing redundant trajectories into lightweight Space-Time Indicators, thereby preserving critical historical information and enhancing spatio-temporal perception. Finally, we design an efficient Two-Level Alignment Paradigm (Align) via Group Relative Policy Optimization (GRPO). By superimposing global outcome rewards with fine-grained process rewards, this dense supervision tightly aligns the agent's cognitive planning with physical environmental feedback, endowing the model with adaptive reasoning capabilities. Experiments demonstrate that TAMP-Nav achieves state-of-the-art performance (e.g., 66.2% SR on R2R-CE) with high runtime and sample efficiency (requiring only 90k training trajectories).

cs.RO↗
arXiv2026

An Asynchronous Triggered MAC Protocol for Underwater Acoustic Networks

Time Division Multiple Access (TDMA)-based Medium Access Control (MAC) protocols have proven their practicality through extensive field trials in Underwater Acoustic Networks (UANs), attributable to their hardware compatibility and ease of implementation. In conventional TDMA-based MAC designs, channel access is typically organized using synchronized, fixed-length slots to mitigate contention and coordinate transmissions. However, this paradigm imposes significant clock synchronization overhead in UANs with long and variable propagation delays and struggles to improve scheduling flexibility. Although some protocols attempt to refine this slot paradigm (adjust the slot length to improve channel reuse efficiency or scheduling frequency), they are still constrained by the trade-off between channel utilization and scheduling complexity. To this end, this paper proposes AT-MAC, an Asynchronous Triggered MAC protocol that aims to achieve efficient and fair channel access through coordinated asynchronous scheduling. AT-MAC introduces a triggered slot paradigm without time synchronization, decoupling transmission scheduling from a rigid timeline and enabling asynchronous, variable-length slots to accommodate the long and diverse propagation delays. To power this slot paradigm, AT-MAC augments conventional Multi-Agent Deep Reinforcement Learning to handle asynchronous interaction, achieving coordinated channel access under partial observations. It further devises a load-aware fairness guard mechanism to enable network-wide fairness status inference solely through local overhearing, thereby guiding adaptive scheduling correction to maintain fairness. Field-reconstructed simulations and on-board inference benchmarking demonstrate the feasibility of AT-MAC. Extensive simulation results further demonstrate its consistent performance gains across the evaluated scenarios and traffic conditions.

cs.NI↗
arXiv2026

Design and Implementation of Schwarz Information Criterion-Aided Intelligent Decentralized Resource Allocation in Dynamic LoRa Networks

This paper proposes a lightweight distributed learning method for selecting transmission parameters in Long-Range (LoRa) networks that adapts to dynamically changing communication environments. In the proposed method, the Thompson Sampling (TS) is adopted for transmission parameter selection, whereas the Schwarz Information Criterion (SIC) is employed for environmental change detection. TS is a reinforcement learning approach that effectively balances exploration and exploitation by updating parameters based on probability distributions. Additionally, it demonstrates stable performance even with a small number of trials, thereby making it well-suited for LoRa end devices (EDs) with limited memory capacity and computational resources. Furthermore, to address the issue that TS-based methods strongly depend on past learning histories and therefore adapt slowly to abrupt changes in communication environments, a statistical change detection mechanism based on the SIC is integrated into our proposed method. SIC is adopted because it can detect environmental changes with low computational cost and is suitable for implementation on resource-constrained LoRa EDs. When a change in the communication environment is detected by SIC, the learning history of TS is reset, thereby enabling rapid re-learning under new environmental conditions. Moreover, to achieve fully distributed communication parameter selection while enhancing transmission reliability and energy efficiency, the proposed method relies solely on Acknowledgment (ACK) feedback and the selected transmission parameters. Experimental results demonstrate that the proposed method improves the transmission success rate from 64.0% to 71.1% and increases energy efficiency from 293.9 bit/J to 328.3 bit/J compared with the conventional Upper Confidence Bound (UCB)1-tuned scheme under high-density dynamic LoRa networks.

cs.DC↗
arXiv2026

Show, Don't Tell: Evaluating Spatial Cognition in Generative Pixels Rather Than LLM Text

Spatial intelligence is essential for agents to move from static semantic understanding toward interacting with the physical world. Many spatial tasks are grounded in continuous visual scenes, where locations, regions, and paths are more naturally expressed by pointing, marking, or drawing than by reporting precise coordinates or discrete textual symbols. Yet existing spatial reasoning benchmarks usually require coordinates, options, or text, creating an answer-interface mismatch for image-generation models. This makes it difficult to evaluate image-generation models under the same task semantics as text-output VLMs, despite their ability to externalize spatial judgments directly in pixel space. We propose ProVisE (Protocolized Visual Evaluation), a benchmark-agnostic framework that elicits protocol-constrained visual answers from image-generation models and parses them into structured predictions compatible with original metrics. ProVisE also includes an Agentic builder that constructs and validates task-specific protocols for new benchmarks. We further introduce SpatialGen-Bench, a curated diagnostic benchmark of 470 samples across 14 spatial subtasks, four capability levels, and diverse answer forms. We evaluate representative text-output VLMs and image-generation models in a unified setting and validate Agentic protocol construction on six external spatial benchmarks. Results show that image-generation models are competitive when spatial answers can be externalized directly in pixel space, while text-output VLMs retain a clear advantage in compositional spatial reasoning. These findings reveal complementary strengths of pixel-space expression and text-based reasoning and establish a metric-compatible testbed for studying spatial cognition in image-generation models.

cs.CV↗
arXiv2026

VLA-Corrector: Lightweight Detect-and-Correct Inference for Adaptive Action Horizon

Vision-Language-Action (VLA) foundation models have recently achieved strong progress in embodied intelligence. To reduce policy-call frequency while preserving temporal coherence, most generative policies adopt an action chunk mechanism, executing multiple future actions in an open-loop manner under a fixed action horizon. However, this "predict-then-blindly-execute" paradigm sacrifices closed-loop reactivity: in contact-rich physical interactions, even small local perturbations can rapidly amplify within the open-loop blind spot, leading to compounding errors and ultimately task failure. To address this limitation, we propose VLA-Corrector, a lightweight corrective inference framework for action-chunked VLA policies. Without modifying the backbone policy weights, VLA-Corrector introduces a lightweight Latent-space Vision Monitor (LVM) that continuously compares predicted and actual visual feature evolution, enabling online detection of visual dynamics deviations. Once persistent deviation is detected, the system triggers a truncation event, discards the remaining stale actions, and invokes corrective replanning via Online Gradient Guidance (OGG). The detect-and-correct mechanism of VLA-Corrector naturally induces an event-triggered adaptive action horizon: it preserves long-horizon execution when the current chunk remains reliable, and invokes short-horizon corrective replanning when execution begins to drift. In doing so, VLA-Corrector mitigates the trade-off imposed by static horizons between execution robustness and policy-call frequency. It can be integrated into different VLA models without further retraining the VLA backbone, interrupting compounding errors while preserving much of the efficiency benefit of action chunking and substantially improving robustness in long-horizon, contact-rich robotic manipulation tasks.

cs.RO↗
arXiv2026

QSplitFL: Capability Aware Deep Q-Learning for Optimal Split Point Selection in Split Federated Learning

Federated Learning (FL) combined with Split Learning (SL) is a privacy preserving paradigm that enables training deep neural networks (DNNs) on resource constrained devices while reducing overall training cost. However, determining the optimal split point, meaning the layer where the model is divided still remains a critical challenge, especially when clients have heterogeneous hardware capabilities. Fixed split points can overload weak devices and increase the communication and server load, which slows convergence and reduces stability. This paper introduces QSplitFL, a novel capability-aware Deep Q-Network (DQN) framework for optimal split point selection in Split learning based Federated Learning (SFL) environments. Unlike existing approaches that rely on high-dimensional model weight representations, QSplitFL employs a lightweight state representation derived directly from client hardware metrics, including CPU utilization, memory, battery level, and network latency. The proposed framework incorporates a decayed loss-drop reward function that prioritizes early convergence, and a committee-based DQN architecture with majority voting to mitigate reward hacking. Extensive experiments on MNIST, Fashion-MNIST, CIFAR-10, and CIFAR-100 datasets using CNN, ResNet50, MobileNetV4, and ConvNeXt architectures demonstrate that our approach achieves better convergence and higher accuracy compared to existing methods, while effectively adapting to heterogeneous device resources. The source code is publicly available at https://github.com/AIPO-Lab/QSplitFL.

cs.LG↗
arXiv2026

Mitigating Manifold Departure: Uncertainty-Aware Subspace Rectification for Trustworthy MLLM Decoding

MLLMs frequently hallucinate objects inconsistent with visual inputs. This issue is typically attributed to the over-reliance on language priors, which can override the visual context. Recent training-free decoding strategies address this by penalizing language priors. However, these methods overlook the dual nature of language priors, where they can be both helpful and harmful depending on the alignment with visual evidence. In particular, blindly suppressing language priors often disrupts the model's semantic manifold, leading to performance degradation, a phenomenon we term Manifold Departure. To address this, we propose Manifold-Guided Adaptive Projection (MGAP), a geometry-aware, training-free decoding method that mitigates hallucinations while preserving representation structure. MGAP first constructs a language-prior subspace from blind hidden states via SVD. During decoding, MGAP projects each multimodal hidden state onto this subspace and applies a consistency-aware gate to adaptively attenuate only the projected prior component, yielding a subspace-selective update that largely preserves the orthogonal semantic components. Extensive experiments on POPE and CHAIR show that MGAP outperforms prior decoding baselines, achieving stronger hallucination suppression without sacrificing coherence.

cs.LG↗
arXiv2026

Audio Pirates: Black-box Audio Watermark Removal via Diffusion Priors

With the rise of AI-generated audio, watermarking has become widely used for detecting misuse and protecting intellectual property. However, adversaries may try to remove these watermarks, making it critical to evaluate how well watermarking schemes withstand removal attacks. Existing attacks are often impractical: they either noticeably degrade perceptual quality or require access to the watermarking scheme. We propose DiffErase, a black-box watermark removal attack that assumes no knowledge of the target watermarking scheme while maintaining perceptual quality. DiffErase perturbs watermarked audio to an intermediate diffusion noise level and regenerates it using a pretrained denoising model, effectively suppressing watermark signals. Theoretical analysis and extensive experiments demonstrate that inaudible audio watermarks are highly vulnerable: across multiple audio domains, DiffErase consistently removes watermarks while preserving perceptual quality. These findings highlight the need for future audio watermarking designs to consider diffusion-based threats. Code and demos are available at https://differase.github.io/DiffErase/.

cs.CR↗
arXiv2026

Distributed 3D Leader-Follower Formation Control with Field-of-View Safety via Control Barrier Functions

This letter proposes a distributed 3D leader-follower formation (3D-LFF) control framework for multi-UAV systems that achieves formation tracking while enforcing perception safety constraints. Maintaining safe, vision-based 3D-LFF is challenging because onboard cameras impose strict Field-of-View (FOV) limitations, and demanding formation commands can drive the leader outside the follower's camera frustum, resulting in loss of visibility. To address this issue, we develop a perception-aware safe control architecture that guarantees visibility by construction. First, we derive a relative kinematic model in a line-of-sight coordinate representation and design a distributed 3D-LFF tracking controller using only locally available relative states. Next, we embed the nominal formation controller within a Control Barrier Function-based Quadratic Program (CBF-QP) safety filter that minimally modifies the commanded velocities to maintain the leader inside the follower's camera frustum while preserving formation tracking whenever feasible. Gazebo simulations and Crazyflie hardware experiments validate the proposed approach, demonstrating accurate formation tracking and effective FOV enforcement, including scenarios in which the nominal desired formation conflicts with visibility constraints.

eess.SY↗
arXiv2026

Is DRL-based MAC Ready for Underwater Acoustic Networks? Exploring Its Practicality in Real Field Experiments

Medium Access Control (MAC) protocols rely on neighbor and environment information to design collision-free access rules for Underwater Acoustic Networks (UANs). Acquiring this information suffers from high communication overhead due to the unique underwater acoustic channel characteristics, such as long propagation delay, spatiotemporal variations in communication quality, and high attenuation. Deep Reinforcement Learning (DRL) is promising to circumvent the UANs' physical constraints and provide a low-overhead solution for underwater MAC protocols, since it can decide access rules based on real-time observation without extra information exchange. However, the unique underwater acoustic channel characteristics impose significant challenges on observation acquisition, training time, and the balance of multiple reward factors for DRL-based MAC protocols. Most existing methods remain at the theoretical level: (1) they design partial intelligent agents failing to achieve fully autonomous access; (2) they assume unreasonable simulation scenarios, weakening the effects of underwater acoustic channel characteristics on MAC protocols. To enhance the practicality of DRL-based MAC protocols, we first analyze the application challenges of DRL in UANs through real field experiments. Based on the above challenges, we propose a DRL-based MAC protocol that considers observation loss and balances multiple reward factors to achieve efficient Entire Autonomous access in the UAN (EA-MAC). To further explore the feasibility of DRL-based MAC protocols, we implement EA-MAC and other state-of-the-art protocols on underwater acoustic modems and evaluate their performance in real field experiments. Experimental results demonstrate that EA-MAC can adaptively determine the scheduling sequence for each node, enabling high-throughput and fair communication in a straightforward manner for UANs.

cs.NI↗
arXiv2026

GFT: From Imitation to Reward Fine-Tuning with Unbiased Group Advantages and Dynamic Coefficient Rectification

Large language models are typically post-trained using supervised fine-tuning (SFT) and reinforcement learning (RL), yet effectively unifying efficient knowledge injection with robust generalization remains challenging. In this work, we provide a training-dynamics analysis showing that SFT can be interpreted as a special case of policy gradient optimization with an extremely sparse implicit reward and unstable inverse-probability weighting, which together lead to single-path dependency, entropy collapse, and gradient explosion. Motivated by this diagnosis, we propose Group Fine-Tuning (GFT), a unified post-training framework that addresses these intrinsic limitations through two mechanisms: Group Advantage Learning, which constructs diverse response groups and derives normalized contrastive supervision to alleviate reward sparsity, and Dynamic Coefficient Rectification, which adaptively bounds inverse-probability weights to stabilize optimization while preserving efficient knowledge injection. Experiments demonstrate that GFT consistently surpasses SFT-based methods and yields policies that integrate more smoothly with subsequent RL training.

cs.AI↗
arXiv2026

SARE: Sample-wise Adaptive Reasoning for Training-free Fine-grained Visual Recognition

Recent advances in Large Vision-Language Models (LVLMs) have enabled training-free Fine-Grained Visual Recognition (FGVR). However, effectively exploiting LVLMs for FGVR remains challenging due to the inherent visual ambiguity of subordinate-level categories. Existing methods predominantly adopt either retrieval-oriented or reasoning-oriented paradigms to tackle this challenge, but both are constrained by two fundamental limitations:(1) They apply the same inference pipeline to all samples without accounting for uneven recognition difficulty, thereby leading to suboptimal accuracy and efficiency; (2) The lack of mechanisms to consolidate and reuse error-specific experience causes repeated failures on similar challenging cases. To address these limitations, we propose SARE, a Sample-wise Adaptive textbfREasoning framework for training-free FGVR. Specifically, SARE adopts a cascaded design that combines fast candidate retrieval with fine-grained reasoning, invoking the latter only when necessary. In the reasoning process, SARE incorporates a self-reflective experience mechanism that leverages past failures to provide transferable discriminative guidance during inference, without any parameter updates. Extensive experiments across 14 datasets substantiate that SARE achieves state-of-the-art performance while substantially reducing computational overhead.

cs.CV↗
arXiv2026

WebWeaver: Breaking Topology Confidentiality in LLM Multi-Agent Systems with Stealthy Context-Based Inference

Communication topology is a critical factor in the utility and safety of LLM-based multi-agent systems (LLM-MAS), making it a high-value intellectual property (IP) whose confidentiality remains insufficiently studied. Existing topology inference attempts rely on impractical assumptions, including control over the administrative agent and direct identity queries via jailbreaks, which are easily defeated by basic keyword-based defenses. As a result, prior analyses fail to capture the real-world threat of such attacks. To bridge this realism gap, we propose \textit{WebWeaver}, an attack framework that infers the complete LLM-MAS topology by compromising only a single arbitrary agent instead of the administrative agent. Unlike prior approaches, WebWeaver relies solely on agent contexts rather than agent IDs, enabling significantly stealthier inference. WebWeaver further introduces a new covert jailbreak-based mechanism and a novel fully jailbreak-free diffusion design to handle cases where jailbreaks fail. Additionally, we address a key challenge in diffusion-based inference by proposing a masking strategy that preserves known topology during diffusion, with theoretical guarantees of correctness. Extensive experiments show that WebWeaver substantially outperforms state-of-the-art (SOTA) baselines, achieving about 60\% higher inference accuracy under active defenses with negligible overhead.

cs.CR↗
arXiv2026

pFedNavi: Structure-Aware Personalized Federated Vision-Language Navigation for Embodied AI

Vision-Language Navigation VLN requires large-scale trajectory instruction data from private indoor environments, raising significant privacy concerns. Federated Learning FL mitigates this by keeping data on-device, but vanilla FL struggles under VLNs' extreme cross-client heterogeneity in environments and instruction styles, making a single global model suboptimal. This paper proposes pFedNavi, a structure-aware and dynamically adaptive personalized federated learning framework tailored for VLN. Our key idea is to personalize where it matters: pFedNavi adaptively identifies client-specific layers via layer-wise mixing coefficients, and performs fine-grained parameter fusion on the selected components (e.g., the encoder-decoder projection and environment-sensitive decoder layers) to balance global knowledge sharing with local specialization. We evaluate pFedNavi on two standard VLN benchmarks, R2R and RxR, using both ResNet and CLIP visual representations. Across all metrics, pFedNavi consistently outperforms the FedAvg-based VLN baseline, achieving up to 7.5% improvement in navigation success rate and up to 7.8% gain in trajectory fidelity, while converging 1.38x faster under non-IID conditions.

cs.CV↗
arXiv2026

Ground What You See: Hallucination-Resistant MLLMs via Caption Feedback, Diversity-Aware Sampling, and Conflict Regularization

While Multimodal Large Language Models (MLLMs) have achieved remarkable success across diverse tasks, their practical deployment is severely hindered by hallucination issues, which become particularly acute during Reinforcement Learning (RL) optimization. This paper systematically analyzes the root causes of hallucinations in MLLMs under RL training, identifying three critical factors: (1) an over-reliance on chained visual reasoning, where inaccurate initial descriptions or redundant information anchor subsequent inferences to incorrect premises; (2) insufficient exploration diversity during policy optimization, leading the model to generate overly confident but erroneous outputs; and (3) destructive conflicts between training samples, where Neural Tangent Kernel (NTK) similarity causes false associations and unstable parameter updates. To address these challenges, we propose a comprehensive framework comprising three core modules. First, we enhance visual localization by introducing dedicated planning and captioning stages before the reasoning phase, employing a quality-based caption reward to ensure accurate initial anchoring. Second, to improve exploration, we categorize samples based on the mean and variance of their reward distributions, prioritizing samples with high variance to focus the model on diverse and informative data. Finally, to mitigate sample interference, we regulate NTK similarity by grouping sample pairs and applying an InfoNCE loss to push overly similar pairs apart and pull dissimilar ones closer, thereby guiding gradient interactions toward a balanced range. Experimental results demonstrate that our proposed method significantly reduces hallucination rates and effectively enhances the inference accuracy of MLLMs.

cs.CV↗
arXiv2025

Schwarz Information Criterion Aided MAB for Resource Allocation in Dynamic LoRa System

This paper proposes a lightweight distributed learning method for transmission parameter selection in Long Range (LoRa) networks that can adapt to dynamic communication environments. In the proposed method, each LoRa End Device (ED) employs the Upper Confidence Bound (UCB)1-tuned algorithm to select transmission parameters including channel, transmission power, and bandwidth. The transmission parameters are selected based on the ACKnowledgment (ACK) feedback returned from the gateway after each transmission and the corresponding transmission energy consumption. Hence, it enables devices to simultaneously optimize transmission success rate and energy efficiency in a fully distributed manner. However, although UCB1-tuned based method is effective under stationary conditions, it suffers from slow adaptation in dynamic environments due to its strong reliance on historical observations. To address this limitation, we integrate the Schwarz Information Criterion (SIC) to our proposed method. SIC is adopted because it enables low-cost detection of changes in the communication environment, making it suitable for implementation on resource-constrained LoRa EDs. When a change is detected by SIC, the learning history of UCB1-tuned is reset, allowing rapid re-learning under the new conditions. Experimental results using real LoRa devices demonstrate that the proposed method achieves superior transmission success rate, energy efficiency, and adaptability compared with the conventional UCB1-tuned algorithm without SIC.

cs.NI↗