arXiv subjects

Song Wang

Publications and source records attributed to Song Wang.

At least 19 recordsLinked to original sources

arXiv2026

Physics-Informed Structure Anchoring With Capture-Aware Prototype Calibration for Cross-Environment RF Fingerprinting

Radio frequency fingerprint identification (RFFI) exploits transmitter-specific hardware imperfections as physicallayer identity cues for Internet of Things (IoT) devices, but deep models often degrade across acquisition environments. In multi-antenna reception, antenna topology and frequencyoffset dynamics structure receiver observations, while capturedependent variation distorts target embeddings and misaligns source-trained decision boundaries. This article proposes physicsinformed structure anchoring with capture-aware prototype calibration (PISA-CAPC) to address both representation and decision mismatches. The two stages separate source representation construction from target decision correction. During source training, PISA organizes antenna tokens through a topology-guided graph, conditions propagation on CFO-derived acquisition dynamics, and applies bounded contextual residual suppression to preserve identity evidence. At deployment, unlabeled capture-aware prototype calibration (U-CAPC) estimates capture-local prototypes and recalibrates target decision scores while keeping the representation and source classifier fixed. Thus, calibration uses neither target labels nor target-domain backbone updates. On a measured WiFi benchmark with four receive antennas and ten transmitters, PISA-CAPC achieves a mean target-domain Macro-F1 of 0.9257 under a balanced transductive setting. Component ablations support complementary roles for topology-guided anchoring, CFO-conditioned modulation, reliability-aware token aggregation, contextual suppression, and capture-aware calibration. These results indicate that physically motivated representation learning can be combined with labelfree decision calibration to improve cross-environment RFFI under the evaluated protocol without changing the deployed backbone.

eess.SP↗

arXiv2026

WorldSR: Harnessing World Knowledge Search for Grounded Image Super-Resolution

Generative-prior-based super-resolution (SR) methods leverage implicit knowledge learned from large-scale models to enhance low-resolution inputs. However, such priors are inherently stochastic and optimized for diverse generation, conflicting with the deterministic, high-fidelity reconstruction required for SR. It is also tied to the base model, costly to update, and prone to hallucinations as real-world data evolves. To address these limitations, we introduce WorldSR, a novel framework that augments SR with explicit, entity-aligned visual world knowledge. WorldSR automatically decomposes the input into semantic entities and performs entity-wise retrieval of structurally and semantically aligned images from a large-scale external corpus. For efficiency and reliability, we incorporate a memory mechanism to reuse previously retrieved results and a multi-step filtering process to refine candidate references. The selected references as world knowledge are then integrated through a novel multi-reference super-resolution model, which provides fine-grained constraints for detail synthesis and structural recovery. Experiments on real-world SR benchmarks show that WorldSR consistently improves structural fidelity and visual quality, outperforming existing generative super-resolution methods. Our results suggest that entity-level grounding in external visual world knowledge provides a simple and effective complement to model-bound generative priors.

cs.CV↗

arXiv2026

MuteBench: Modality Unavailability Tolerance Evaluation for Incomplete Multimodal Fusion

Multimodal physiological data powers clinical AI systems from intensive care units to wearable devices, but sensors routinely fail in practice. Two failure modes are common: modality missing, where an entire channel is absent, and within-modality missing, where a contiguous time segment is lost. No existing benchmark evaluates multiple fusion architectures under both failure modes at controlled severity levels across diverse clinical datasets. We present MuteBench, a benchmark covering 9 datasets from 7 clinical domains, 6 fusion architectures, and 2 missing-data modes over 125,000 samples. Through this benchmark, we find that architecture family is the strongest predictor of robustness, outweighing parameter count. Channel-independent models tolerate modality missing well but can be sensitive to within-modality missing, especially on short sequences. Curriculum modality dropout protects reliably only up to the maximum dropout rate used in training. We also find that channel count, sequence length, and modality alignment jointly determine which failure mode poses the greater threat. Finally, a PTB-XL case study suggests that diffusion-based imputation can improve downstream classification under within-modality missing, with the largest gains for models whose expert routing is most sensitive to corrupted inputs, though broader validation across datasets remains an open direction. MuteBench provides practitioners with concrete guidance for both selecting existing architectures and informing the design of future robust multimodal fusion methods.

cs.LG↗

arXiv2026

Better Specifications, Better Code: Fine-Tuning LLMs for Behavioral Contract Generation

Large language models (LLMs) demonstrate strong code-generation capabilities but often struggle with functional correctness due to implicit or ambiguous behavioral requirements. We present CG2, a contract-guided code generation framework that translates natural-language tasks into structured behavioral contracts specifying interfaces, preconditions, postconditions, and invariants. CG2 fine-tunes an LLM to generate these contracts and introduces an iterative semantic validation and repair process to identify and correct incomplete, inconsistent, or incorrect specifications before guiding an unchanged downstream code-generation model. We evaluate CG2 on 1,007 held-out tasks using five code LLMs: DeepSeek-Coder-V2-Lite-Instruct, Ministral-3-14B-Instruct-2512, Qwen2.5-Coder-14B-Instruct, Qwen2.5-Coder-32B-Instruct, and Qwen3.5-27B. Results show that CG2 consistently improves pass@1 over direct code generation by 1.49-3.27 percentage points (7.58%-22.45% relative improvement) and over contract guidance without fine-tuning by 1.19-2.38 percentage points (7.04%-14.65% relative improvement). Manual analysis further reveals that semantically correct contracts are associated with higher code-generation accuracy. These findings demonstrate that explicitly specifying and validating behavioral requirements can improve functional correctness without modifying downstream code-generation models.

cs.SE↗

arXiv2026

How Effectively Do Security Skills Guide Coding Agents in Vulnerability Detection in Linux Kernel?

Large language model (LLM) coding agents, such as Claude Code, are increasingly used for security auditing, with task-specific skills providing reusable vulnerability-analysis procedures. However, their effectiveness in detecting real vulnerabilities and the reasons for their failures remain unclear. We conduct the first large-scale empirical study of 37 security skills using Claude Code with four underlying LLMs on 100 real Linux kernel vulnerabilities across ten classes, totaling 14800 runs. We evaluate vulnerability detection at the function, statement, and vulnerability-class levels. Results show that 24.4 percent of runs locate the correct function, 18.3 percent identify both the function and responsible statement, and 15.7 percent correctly classify the vulnerability. Our failure analysis reveals limitations in evidence acquisition, state-relation reasoning, code interpretation, candidate verification, investigation coverage, and reporting. Based on these findings, we propose SKILL-SIF, a self-improving framework that iteratively evolves security skills by learning from missed vulnerabilities and failure patterns. We evolve five top-performing skills using the original benchmark and evaluate them on 100 previously unseen vulnerabilities. SKILL-SIF improves fully correct vulnerability detection by up to eight times (from 2 percent to 16 percent) on unseen vulnerabilities, while aggregate accuracy across four LLMs increases by 1.85 times (from 6.05 percent to 11.20 percent). These findings highlight the limitations of existing security skills and demonstrate the potential of failure-driven skill evolution to improve vulnerability detection in coding agents.

cs.CR↗

arXiv2026

SOMA: Continuous Monitoring and Intervention for System Two Reasoning with System One Models

Reasoning models often run into problems without noticing them, even when symptoms are visible in their own text. In humans, metacognition separates monitoring from control: fast, intuitive signals keep watch while slow, deliberate reasoning unfolds and help decide whether to continue, change course, or stop. System One (S1) decision models, such as Jev, are a natural candidate for this role: they return a calibrated probability for a natural-language question about a text in one fast call, yet they are used only as gates before reasoning. We introduce SOMA (System One Metacognitive Awareness), in which an S1 model continuously monitors a System Two (S2) reasoning model and intervenes when needed. SOMA has three layers: a scenario layer states each objective in natural language with predefined actions, an awareness layer lets S1 judge whether the described behavior appears, and a control layer accumulates these judgments, first reminds S2, and takes over only if the behavior persists. Sharing one S1 model and one controller across tasks, without training, SOMA addresses two different failures on two reasoning models: on overthinking, it cuts reasoning tokens by up to 36\% while improving average accuracy, and on unanswerable questions, it raises abstention from 70.5\% to 87.2\%.

cs.AI↗

arXiv2026

PROVISO: Provenance-Informed Action Selection for Secure LLM Agents

LLM agents act on the external world through tools and third-party skills, so a single executed action can cause irreversible harm. Existing defenses judge a proposed action without tracking whether its content comes from the user or from external sources such as tool outputs, and they judge it in isolation rather than against an available alternative, which forces a trade-off between security and utility. We present PROVISO, a controller that decides which action to commit after the agent proposes an action and before it is executed. PROVISO encodes trusted observations, external observations, the proposed action, and an alternative into provenance-typed memory, relates them through role-specific queries, and commits the action with the highest calibrated value learned from outcomes. On four agent security tasks spanning the agent lifecycle, no baseline exceeds PROVISO in both security and utility. PROVISO improves F1 over the strongest baseline by 13.84 points on SkillsBench and by 7.07 points on ToolSafe, and it reaches 88-99 F1 on three additional skill detection benchmarks.

cs.AI↗

arXiv2026

PORA: Learning When More Reasoning Pays Off for LLM Agents

LLM agents perform tasks over multiple steps and revise their actions based on feedback. As the total reasoning budget is often limited and must support future steps, the agent must choose between short and long reasoning before each action. Longer reasoning may improve the final task outcome by preventing later errors, but it also uses budget capacity that may be needed for future decisions. This creates two challenges: estimating the downstream value of long reasoning and deciding when that value is worth using the remaining budget. We propose Paired Outcome-guided Reasoning Allocation (PORA), which learns this decision by comparing short and long reasoning from the same interaction state. During training, \method{} completes the task twice from each selected state, once with a short reasoning allowance and once with a long allowance, and compares their final quality and total processing to estimate when longer reasoning is valuable. During inference, a controller weighs this estimate against learned prices on remaining-budget pressure and chooses either short or long reasoning without running both alternatives. We evaluate PORA across application, memory, code, retail, and mathematics tasks, achieving strong results across all settings. On BFCL memory, it exceeds the strongest existing method by 17.4--32.9 percentage points. Within the same agent, it matches or exceeds always using long reasoning on code and mathematics with 50.6% fewer actor tokens and 31.2% fewer model requests, respectively.

cs.AI↗

arXiv2026

PatientWM: Towards Generalizable Clinical World Models via Unified Patient States

Clinical decisions depend on anticipating how a patient's condition will evolve as treatment is adjusted and new evidence becomes available. Existing medical world models capture patient-state evolution, but often define the evolving state around a particular modality or task, limiting reuse as inputs or prediction targets change. These models also retain history through observation sequences or recurrent or latent summaries rather than exposing preceding patient states and clinical actions as an ordered transition context. This distinction matters because similar current measurements may follow different recent changes, treatment exposures, and observed responses. To tackle these limitations, we propose PatientWM, a clinical world-model framework built around a unified patient state and sequential latent history. Availability-aware encoders integrate observations available at each decision time into the patient state, while typed decoders produce task-specific clinical readouts without defining the state around any single target. PatientWM further conditions each next-state prediction on the current state and supplied action together with a bounded, ordered history of preceding latent patient states and clinical actions, preserving recent treatment and response context. Experiments across MIMIC-III, MIMIC-IV, and three external clinical datasets show superior performance in clinical prediction, multi-step forecasting, and action-planning evaluation. Transfer experiments further demonstrate the generalizability of PatientWM's patient state. Together, these results establish its ability to forecast clinical evolution and support planning.

cs.AI↗

arXiv2026

Quantization-Robust Unlearning through the Lens of Retain-Forget Loss Landscapes Interaction

Unlearning ensures LLM compliance by removing the influence of private or copyrighted training data. However, since LLM models typically undergo post-training compression, like quantization, in practical deployment, it has been observed that the unlearning effect can be substantially weakened, with the forgetting behavior degrading more severely than that of model utility. This paper proposes a quantization-robust unlearning framework that makes forgetting robust to quantization while maintaining overall model utility. We analyze this gap through the lens of loss landscape. Specifically, our analysis reveals a curvature-based criteria that pinpoints sensitive weights in the unlearned model that leads to both non-robust forgetting and reduced utility. We therefore propose sensitivity-guided noisy regularization, which is applied on the sensitive parameters to steer the model convergence towards a smoother minima of uniformly low forget and retain losses. Balancing unlearning and utility, we further propose forget-critical optimization, which updates only forget-critical layers, preserving most of the network to retain useful knowledge. Extensive experiments on the MUSE and TOFU benchmarks across multiple LLM unlearning algorithms show that our approach achieves substantially more quantization-resilient forgetting while maintaining utility. Code is available at https://github.com/jialuw0830/QRUnlearn.

cs.LG↗

arXiv2026

AutoAdapt: Reliable Few-Shot Adaptation under Clinical Distribution Shifts

Large pretrained clinical models provide a practical way to reuse learned prior knowledge across hospitals by adapting models to them. In practice, a target hospital may only have a small labeled patient cohort, a setting commonly referred to as few-shot adaptation. This requires making multiple decisions, such as which pretrained model to adapt, how much of the model to update, and which patients to use. Nevertheless, this process faces two primary challenges. First, the best adaptation strategy varies across clinical tasks. Second, evaluating and comparing candidate strategies becomes unreliable due to the small patient cohort. In this work, we introduce AutoAdapt with two core designs to deal with these challenges. The Adapter defines an extensible space of adaptation recipes, and the Automator forms a weighted recipe combination from evidence within the adaptation patients. We propose a reliability rule to ensure that only the most effective strategy on most available patients will be selected. These selected strategies then form a combination for effective few-shot adaptation. We conduct extensive experiments across critical care, emergency care, and diagnostic datasets, and the results show that AutoAdapt consistently achieves state-of-the-art performance using only a few patients for adaptation.

cs.CV↗

arXiv2026

MCFR: A Mask-Guided Coarse-to-Fine Regression Framework for Robust Multi-Variant Board-to-Board Connector Assembly

Automated insertion of board-to-board (BTB) connectors in 3C manufacturing requires both high visual accuracy and strong deployment robustness. This problem remains challenging because multi-variant connectors exhibit significant morphological and appearance variations, making stable cross-variant generalization difficult, while the mismatch between training and deployment under fixed-view inspection settings induces background spurious correlation and degrades real-world performance. To address these issues, this paper proposes MCFR, a Mask-Guided Coarse-to-Fine Regression framework for multi-variant BTB connector assembly. By introducing an object-aware mask prior and explicit photometric refinement, the proposed method suppresses background interference and improves alignment accuracy and robustness in practical deployment. Experiments on a self-constructed multi-variant dataset, a BTB batch insertion testbed, and a real smartphone assembly task show that MCFR consistently outperforms representative baselines and achieves an average real-world insertion success rate of 99.25%. These results demonstrate the effectiveness and practical potential of MCFR for automated assembly of multi-variant BTB connectors.

cs.RO↗

arXiv2026

TAP: Efficient Long-Horizon Agent Pruning via Trajectory-Anchored Recovery

Emerging long-horizon agentic tasks require repeated model calls, worsening the inference cost of already-costly language models. While narrow agentic tasks suggest potential for aggressive model pruning without performance drop, empirical results show existing methods proposed for question answering tasks severely degrade task performance when applied to agentic models. We trace this failure to two decisions: what to prune and how to recover. For pruning, one-shot importance estimates fail to track how the pruned model adapts. For recovery, offline distillation covers only teacher prefixes, while full-trajectory on-policy distillation causes student errors to compound across turns. In this work, we propose Trajectory-Anchored Pruning (TAP), the first structural pruning framework for reinforcement learning (RL)-trained agents. TAP couples structural pruning with efficient on-policy recovery, anchoring interactions to teacher trajectories while allowing the student to generate each reasoning-action response. A frozen dense teacher supervises the student's response prefixes, addressing within-response training-inference mismatch while preventing student-induced deviations from propagating across training turns. Instead of one-shot pruning, TAP re-scores channels using gradients of the recovery objective on the recovered student, connecting iterative channel selection to the evolving policy. With 60% of FFN channels removed, TAP retains 99.2% and 88.0% of the dense 7B agents' task success rates on ALFWorld and WebShop, respectively, while reducing GPU time per successful task by approximately 22% and 17%. These results demonstrate effective structural compression of long-horizon agents under a limited recovery budget.

cs.LG↗

arXiv2026

VisAudit: Evaluating Multimodal Agents for Visual Diagnosis and Repair

Multimodal agents are increasingly used for data visualization tasks but remain limited in autonomous review. Unlike humans, they may fail to recognize when a visualization is incorrect, determine what to change, repair it without disrupting correct content, and verify whether the intervention succeeded. Existing benchmarks largely evaluate predefined individual capabilities such as chart generation, instruction-guided editing, or defect detection, and therefore do not capture this gap in autonomous review. We introduce VisAudit, a benchmark for evaluating visualization diagnosis, repair, and verification. Given a rendered chart and configurable auxiliary evidence, including its source data table, intended text summary, and visualization code, an agent iteratively diagnoses potential defects, modifies and executes visualization code, inspects execution and visual feedback, and determines when no further intervention is needed. VisAudit defines three tracks spanning diagnosed repair, autonomous repair, and open-world verification, and contains 1,900 flawed instances across 21 chart types and 10 flaw categories, together with 300 initially correct charts. We construct the benchmark through controlled perturbations of validated source visualizations, with systematic verification and human-aligned quality control to ensure that injected defects are well-defined and recoverable from the available evidence. Experiments with leading multimodal models reveal a substantial gap from reliable autonomous review: the strongest evaluated model fully recovers only $47.4\%$ of flawed charts in the autonomous-repair setting.

cs.LG↗

arXiv2026

ReF-HIL: Shaping the Critic around Human Action Neighborhoods for Efficient Human-in-the-Loop Reinforcement Learning

Human-in-the-loop reinforcement learning (HIL-RL) offers a promising route to efficient training of robotic manipulation policies by combining autonomous learning with human demonstrations and online corrections. However, insufficient use of successful human experience in value learning prolongs costly real-world training, while persistent imitation penalties can limit value-driven policy improvement. To address these limitations, we propose ReF-HIL, an efficient HIL-RL framework that uses human guidance to accelerate the learning process. Human-Reference-Guided Value Shaping learns an independent value reference from successful human experience to guide online value learning, while incorporating local corrective feedback. A Human Action Fence defines a learned human-action neighborhood, allowing value-driven optimization for better performance without imitation penalties inside while constraining policy and value updates outside. Experiments on five diverse and challenging real-world manipulation tasks demonstrate improved overall learning efficiency and higher success rates compared with the evaluated baselines. Specifically, ReF-HIL reaches 90% autonomous success in only 18-63 minutes of active training and achieves final success rates of 91.7-100%. These results highlight the potential of human-guided reinforcement learning to acquire reliable manipulation skills efficiently in the real world. Project website: https://anonymous.4open.science/w/ReF-HIL-7762/

cs.RO↗

arXiv2026

PrivMeSA: Privacy-Aware Self-Evolving Multi-Agent System for Medicine via Local-Remote LLM Collaboration

Clinical large language model (LLM) agents deployed locally can consult more capable remote models, but doing so risks exposing patient information. Privacy-conscious delegation places disclosure decisions with a local agent, yet removing explicit identifiers is insufficient: quasi-identifiers can accumulate across multi-turn consultations and repeated patient visits to enable re-identification. We introduce PrivMeSA, a privacy-aware self-evolving multi-agent system that learns to control disclosure and retains remote expertise for local reuse. A local agent manages each encounter and consults remote specialists that may request additional information. Reinforcement learning balances task accuracy against direct disclosure and registry-based re-identification risk, with privacy evaluated over the complete outbound transcript of each encounter. A local lesson memory distills completed consultations into generalized clinical guidance and retrieves relevant lessons before transmission, allowing subsequent cases to reuse expertise without another remote exchange. Memory grows without additional outcome labels or parameter updates. On an emergency-department benchmark built from MIMIC-IV-ED records, PrivMeSA improves mean task accuracy over delegation by up to 15.8 percentage points. In the same setting, PrivMeSA reduces the disclosure of personal details from 98.0% to 0.2% of cases and the share of cases in which the patient can be narrowed to ten or fewer registry patients from 74% to 0%.

cs.AI↗

arXiv2026

MLLMs Hallucinate when Information Distribution Drifts in Synergy Heads

Multimodal Large Language Models (MLLMs) often struggle with hallucinations, thus hindering their reliable practical applications. Existing attention-based mitigation methods mainly rely on indirect signals (e.g., attention weights) that fail to accurately reflect the actual information shift underlying hallucination generation. In this paper, we propose HEAL, Head-lEvel information disentAnglement and caLibration for identifying and mitigating hallucinations. HEAL first employs causal noise intervention on multi-head outputs to filter out causally redundant heads. Subsequently, it disentangles information distribution within the remaining heads via the counterfactual Difference-in-Differences, categorizing heads into four types. Through analysis, we observe: hallucinations happen when information distribution drifts away from a healthy equilibrium in synergy heads, not strongly correlated with the quantity or strength of modality-specific heads. Motivated by this insight, HEAL injects dynamic information calibration factors into the value vectors of synergy heads and actively regulates visual-language dependencies, steering the output distribution towards factual evidence. Extensive experiments demonstrate that HEAL effectively reduces hallucinations across multiple MLLMs, offering a simple and interpretable pathway to enhance model trustworthiness.

cs.CV↗

arXiv2026

Spackle: Completing Large View Single Image NVS with Adaptive Gaussians

Single-image novel view synthesis (NVS) enables photorealistic rendering of un- observed viewpoints from a single input. Practical NVS systems require two key capabilities: robust reconstruction of occluded regions and high inference effi- ciency. While hybrid decoupled frameworks combining feedforward 3D Gaussian Splatting (3DGS) and diffusion models show promise for large-view-deviation NVS, they suffer from capacity competition: a fixed number of Gaussians forces resource shifts from visible to newly disoccluded areas, degrading original scene fidelity when the target view deviates significantly from the input. To address this, we propose Spackle, a lightweight residual learning framework that mit- igates capacity competition without sacrificing efficiency. Spackle operates in three stages: predicting base 3DGS attributes from given views, automatically identifying poorly reconstructed regions, and learning a residual 3DGS optimized exclusively for these areas. At inference, we combine the baseline and aug- mented Gaussians for NVS. We conduct comprehensive experiments and show that Spackle achieves state-of-the-art performance on large-view-deviation cases.

cs.CV↗