Search arXiv⌕ Search

arXiv subjects

Jingyi Wang

Publications and source records attributed to Jingyi Wang.

At least 37 records · Page 2Linked to original sources
arXiv2026

Learning from Your Own Mistakes: Constructing Learnable Micro-Reflective Trajectories for Self-Distillation

Self-distillation improves reasoning in large language models by using the model's own rollouts as training signal, typically through implicit logit-level alignment that minimizes KL divergence toward a privileged target distribution. However, because this supervision is generated via uncontrolled sampling, it provides no diagnostic insight into the model's specific errors or corrective guidance for its individual failure patterns. Consequently, the model learns to imitate a privileged distribution rather than receiving fine-grained corrections that pinpoint where and why its reasoning fails. In this paper, we propose Trajectory-Augmented Policy Optimization (TAPO), which advances self-distillation from implicit distributional alignment to explicit trajectory construction. During RL training, the model produces both correct and incorrect rollouts to the same query, and TAPO leverages this contrastive structure to construct micro-reflective corrections, new training trajectories that retain the model's erroneous reasoning up to the point of failure, then insert a natural-language diagnosis and corrected reasoning guided by a correct reference from the same sampling group. Since each trajectory is anchored in the learner's own prefix and solutions, the corrective signal preserves the model's on-policy distribution to a greater extent than the position-wise alignment imposed by KL-based methods. To integrate these trajectories, TAPO introduces difficulty-aware candidate selection at the model's capability boundary and decoupled advantage estimation to prevent gradient contamination. Experiments on AIME 2024, AIME 2025, and HMMT 2025 show that TAPO achieves consistent improvements over GRPO under the same number of training steps. Further analysis demonstrates that TAPO strengthens both first-pass reasoning and error-correction effectiveness.

cs.LG↗
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

Modified uni-travelling-carrier photodiodes with 206 GHz bandwidth and 0.81 A/W external responsivity

The accelerating demand for wireless communication necessitates wideband, energy-efficient photonic sub-terahertz (sub-THz) sources to enable ultra-fast data transfer. However, as critical components for THz photonic mixing, photodiodes (PDs) face a fundamental trade-off between quantum efficiency and bandwidth, presenting a major obstacle to achieving high-speed performance with high optoelectronic conversion efficiency. Here, we overcome this challenge by demonstrating an InP-based, waveguide-integrated modified uni-traveling carrier photodiode (MUTC-PD) with bandwidth exceeding 200 GHz and a bandwidth-efficiency product (BEP) surpassing 130 GHz. Incorporating a spot-size converter (SSC) together with optimized electric-field distribution, balanced carrier transport, and minimized parasitic capacitance, the device achieves a 3-dB bandwidth of 206 GHz and an external responsivity of 0.81 A/W, setting a new BEP benchmark. Packaged with WR-5.1 waveguide output, it delivers radio-frequency (RF) power exceeding -5 dBm across the 127-185 GHz frequency range. As a proof of concept, we achieved a wireless transmission of 54 meters with a single-line rate of up to 120 Gbps, leveraging photonics-aided technology without requiring a low-noise amplifier (LNA). This work establishes a pathway to significantly enhance optical power budgets and reduce energy consumption, presenting a transformative step toward high-bandwidth, high-efficiency sub-THz communication systems and next-generation wireless networks.

physics.app-ph↗
arXiv2026

Provable Fairness Repair for Deep Neural Networks

Deep neural networks (DNNs) are suffering from ethical issues such as individual discrimination. In response, extensive NN repair techniques have been developed to adjust models and mitigate such undesired behaviors. However, existing fairness repair methods are typically data-centric, which often lack provable guarantees and generalization to unseen samples. To overcome these limitations, we propose ProF, a novel fairness repair framework with provable guarantees. The key intuition of ProF is to leverage interval bound propagation (a widely used NN verification technique) to soundly capture model outputs over the whole set $S(\mathbf{x})$ around a biased sample $\mathbf{x}$. The derived bounds are utilized to guide fairness repair which encourages the model to produce consistent outputs on $S(\mathbf{x})$. Specifically, we integrate fairness constraints and model modifications into a unified constraint-solving formulation, which can be transformed to a Mixed-Integer Linear Programming (MILP) problem solvable by off-the-shelf solvers. The solution to the MILP problem effectively induces a repaired model with guaranteed fairness over the whole set $S(\mathbf{x})$. We evaluate ProF on four widely used benchmark datasets and demonstrate that it achieves provable fairness repair, with generalization of up to 95.93\% on full datasets and 93.16\% on the entire input space. Notably, ProF can be easily configured to support multiple sensitive attributes and more practical fairness definitions, while providing provable repair guarantees and delivering around 90\% fairness improvement. Our code is available at https://github.com/nninjn/ProF.

cs.SE↗
arXiv2026

Tropical Degenerations of Network Games:Valuation Classes and Equilibrium Coalescence

A valuation-theoretic framework is developed for studying tropical degenerations of multilinear network games. Equilibrium conditions are modeled by an ideal over the Puiseux field, and valuation classes and cluster multiplicities are used to describe the organization of Puiseux equilibrium branches under degeneration. For valuation vectors lying in the relative interiors of generator-wise maximal tropical cells, multilinearity is shown to force a binomial reduction of the generator-wise initial system. The resulting binomial systems are governed by exponent-difference graphs, strongly connected component decompositions, and lattice indices computed via Smith normal form. In particular, unimodular diagonal blocks yield initial-coefficient rigidity, whereas non-unimodular blocks give rise to torsion-type leading-coefficient multiplicities. The generic binomial theory is complemented by a collision-normalized cross-prism family. In this family, 2^L Puiseux equilibrium branches share the same valuation vector and the same leading-coefficient vector. The corresponding limiting initial fiber is supported at a single torus point, but this point is nonreduced with scheme-theoretic length 2^L. Thus valuation coalescence is realized as an intrinsic scheme-theoretic collision rather than merely as a loss of higher-order terms. These local degeneration invariants are then related to the algebraic-degree theory of network games, showing how global equilibrium counts can be refined by valuation classes, binomial initial systems, Smith lattice data, and nonreduced collision fibers.

math.AG↗
arXiv2026

GRPO-VPS: Enhancing Group Relative Policy Optimization with Verifiable Process Supervision for Effective Reasoning

Reinforcement Learning with Verifiable Rewards (RLVR) has advanced the reasoning capabilities of Large Language Models (LLMs) by leveraging direct outcome verification instead of learned reward models. Building on this paradigm, Group Relative Policy Optimization (GRPO) eliminates the need for critic models but suffers from indiscriminate credit assignment for intermediate steps, which limits its ability to identify effective reasoning strategies and incurs overthinking. In this work, we introduce a model-free and verifiable process supervision via probing the model's belief in the correct answer throughout its reasoning trajectory. By segmenting the generation into discrete steps and tracking the conditional probability of the correct answer appended at each segment boundary, we efficiently compute interpretable segment-wise progress measurements to refine GRPO's trajectory-level feedback. This approach enables more targeted and sample-efficient policy updates, while avoiding the need for intermediate supervision derived from costly Monte Carlo rollouts or auxiliary models. Experiments on mathematical and general-domain benchmarks show consistent gains over GRPO across diverse models: up to 2.6-point accuracy improvements and 13.7% reasoning-length reductions on math tasks, and up to 2.4 points and 4% on general-domain tasks, demonstrating strong generalization.

cs.LG↗
arXiv2026

Algebraic Multigrid with Filtering: An Efficient Preconditioner for Interior Point Methods in Large-Scale Contact Mechanics Optimization

Large-scale contact mechanics simulations are crucial in many engineering fields such as structural design and manufacturing. In the frictionless case, contact can be modeled by minimizing an energy functional; however, these problems are often nonlinear, nonconvex, and increasingly difficult to solve as mesh resolution increases. In this work, we employ a Newton-based interior-point (IP) filter line-search method, an effective approach for large-scale constrained optimization. While this method converges rapidly, each iteration requires solving a large saddle-point linear system that becomes ill-conditioned as the optimization process converges, largely due to IP treatment of the contact constraints. Such ill-conditioning can hinder solver scalability and increase iteration counts with mesh refinement. To address this, we introduce a novel preconditioner, AMG with filtering (AMGF), tailored to the Schur complement of the saddle-point system. Building on the classical AMG solver, commonly used for elasticity, we augment it with a specialized subspace correction that filters near null space components introduced by contact interface constraints. Through theoretical analysis and numerical experiments on a range of linear and nonlinear contact problems, we demonstrate that AMGF achieves mesh independent convergence and maintains robustness against the ill-conditioning that notoriously plagues IP methods. These results indicate that AMGF makes contact mechanics simulations more tractable and broadens the applicability of Newton-based IP methods in challenging engineering scenarios. More broadly, AMGF is well suited for problems where solver performance is limited by a low-dimensional subspace, such as those arising from localized constraints, interface conditions or model heterogeneities, making it applicable beyond contact mechanics and constrained optimization.

math.NA↗
arXiv2026

Shapley-Guided Neural Repair Approach via Derivative-Free Optimization

DNNs are susceptible to defects like backdoors, adversarial attacks, and unfairness, undermining their reliability. Existing approaches mainly involve retraining, optimization, constraint-solving, or search algorithms. However, most methods rely on gradient calculations, restricting applicability to specific activation functions (e.g., ReLU), or use search algorithms with uninterpretable localization and repair. Furthermore, they often lack generalizability across multiple properties. We propose SHARPEN, integrating interpretable fault localization with a derivative-free optimization strategy. First, SHARPEN introduces a Deep SHAP-based localization strategy quantifying each layer's and neuron's marginal contribution to erroneous outputs. Specifically, a hierarchical coarse-to-fine approach reranks layers by aggregated impact, then locates faulty neurons/filters by analyzing activation divergences between property-violating and benign states. Subsequently, SHARPEN incorporates CMA-ES to repair identified neurons. CMA-ES leverages a covariance matrix to capture variable dependencies, enabling gradient-free search and coordinated adjustments across coupled neurons. By combining interpretable localization with evolutionary optimization, SHARPEN enables derivative-free repair across architectures, being less sensitive to gradient anomalies and hyperparameters. We demonstrate SHARPEN's effectiveness on three repair tasks. Balancing property repair and accuracy preservation, it outperforms baselines in backdoor removal (+10.56%), adversarial mitigation (+5.78%), and unfairness repair (+11.82%). Notably, SHARPEN handles diverse tasks, and its modular design is plug-and-play with different derivative-free optimizers, highlighting its flexibility.

cs.SE↗
arXiv2026

OddGridBench: Exposing the Lack of Fine-Grained Visual Discrepancy Sensitivity in Multimodal Large Language Models

Multimodal large language models (MLLMs) have achieved remarkable performance across a wide range of vision language tasks. However, their ability in low-level visual perception, particularly in detecting fine-grained visual discrepancies, remains underexplored and lacks systematic analysis. In this work, we introduce OddGridBench, a controllable benchmark for evaluating the visual discrepancy sensitivity of MLLMs. OddGridBench comprises over 1,400 grid-based images, where a single element differs from all others by one or multiple visual attributes such as color, size, rotation, or position. Experiments reveal that all evaluated MLLMs, including open-source families such as Qwen3-VL and InternVL3.5, and proprietary systems like Gemini-2.5-Pro and GPT-5, perform far below human levels in visual discrepancy detection. We further propose OddGrid-GRPO, a reinforcement learning framework that integrates curriculum learning and distance-aware reward. By progressively controlling the difficulty of training samples and incorporating spatial proximity constraints into the reward design, OddGrid-GRPO significantly enhances the model's fine-grained visual discrimination ability. We hope OddGridBench and OddGrid-GRPO will lay the groundwork for advancing perceptual grounding and visual discrepancy sensitivity in multimodal intelligence. Code and dataset are available at https://wwwtttjjj.github.io/OddGridBench/.

cs.CV↗
arXiv2026

Practical Efficient Global Optimization is No-regret

Efficient global optimization (EGO) is one of the most widely used noise-free Bayesian optimization algorithms.It comprises the Gaussian process (GP) surrogate model and expected improvement (EI) acquisition function. In practice, when EGO is applied, a scalar matrix of a small positive value (also called a nugget or jitter) is usually added to the covariance matrix of the deterministic GP to improve numerical stability. We refer to this EGO with a positive nugget as the practical EGO. Despite its wide adoption and empirical success, to date, cumulative regret bounds for practical EGO have yet to be established. In this paper, we present for the first time the cumulative regret upper bound of practical EGO. In particular, we show that practical EGO has sublinear cumulative regret bounds and thus is a no-regret algorithm for commonly used kernels including the squared exponential (SE) and Matérn kernels ($ν>\frac{1}{2}$). Moreover, we analyze the effect of the nugget on the regret bound and discuss the theoretical implication on its choice. Numerical experiments are conducted to support and validate our findings.

stat.ML↗
arXiv2026

Programmable electro-optic frequency comb empowers integrated parallel convolution processing

Integrated photonic convolution processors make optical neural networks (ONNs) a transformative solution for artificial intelligence applications such as machine vision. To enhance the parallelism, throughput, and energy efficiency of ONNs, wavelength multiplexing is widely applied. However, it often encounters the challenges of low compactness, limited scalability, and high weight reconstruction latency. Here, we proposed and demonstrated an integrated photonic processing unit with a parallel convolution computing speed of 1.62 trillion operations per second (TOPS) and a weight reconstruction speed exceeding 38 GHz. This processing unit simultaneously achieves, for the first time, multi-wavelength generation and weight mapping via a single programmable electro-optic (EO) frequency comb, featuring unprecedented compactness, device-footprint independent scalability, and near-unity optical power conversion efficiency (conversion efficiency from input optical power to output weighted comb lines). To demonstrate the reconfigurability and functionality of this processing unit, we implemented image edge detection and object classification based on EO combs obtained using the particle swarm algorithm and an EO comb neural network training framework, respectively. Our programmable EO comb-based processing framework establishes a new paradigm towards the development of low-latency monolithic photonic processors, promising real-time in-sensor learning for autonomous vehicles, intelligent robotics, and drones.

physics.optics↗
arXiv2026

Connection Probabilities Estimation in Multi-layer Networks via Iterative Neighborhood Smoothing

Understanding the structural mechanisms of multi-layer networks is essential for analyzing complex systems characterized by multiple interacting layers. This work studies the problem of estimating connection probabilities in multi-layer networks and introduces a new Multi-layer Iterative Connection Probability Estimation (MICE) method. The proposed approach employs an iterative framework that jointly refines inter-layer and intra-layer similarity sets by dynamically updating distance metrics derived from current probability estimates. By leveraging both layer-level and node-level neighborhood information, MICE improves estimation accuracy while preserving computational efficiency. Theoretical analysis establishes the consistency of the estimator and shows that, under mild regularity conditions, the proposed method achieves an optimal convergence rate comparable to that of an oracle estimator. Extensive simulation studies across diverse graphon structures demonstrate the superior performance of MICE relative to existing methods. Empirical evaluations using brain network data from patients with Attention-Deficit/Hyperactivity Disorder (ADHD) and global food and agricultural trade network data further illustrate the robustness and effectiveness of the method in link prediction tasks. Overall, this work provides a theoretically grounded and practically scalable framework for probabilistic modeling and inference in multi-layer network systems.

stat.ME↗
arXiv2026

Attention to details, logits to truth: visual-aware attention and logits enhancement to mitigate hallucinations in LVLMs

Existing Large Vision-Language Models (LVLMs) exhibit insufficient visual attention, leading to hallucinations. To alleviate this problem, some previous studies adjust and amplify visual attention. These methods present a limitation that boosting attention for all visual tokens inevitably increases attention to task irrelevant tokens. To tackle this challenge, we propose a training free attentional intervention algorithm to enhance the attention of task-relevant tokens based on the argument that task-relevant tokens generally demonstrate high visual-textual similarities. Specifically, the vision-text cross-attention submatrices, which represent visual-textual correlations, are extracted to construct the reweighting matrices to reallocate attention. Besides, to enhance the contribution of visual tokens, we inject visual attention values into the beam search decoding to identify solutions with higher visual attention. Extensive experiments demonstrate that this method significantly reduces hallucinations across mainstream LVLMs, while preserving the accuracy and coherence of generated content.

cs.CV↗
arXiv2026

Convergence Rates of Constrained Expected Improvement

Constrained Bayesian optimization (CBO) methods have seen significant success in black-box optimization with constraints. One of the most commonly used CBO methods is the constrained expected improvement (CEI) algorithm. CEI is a natural extension of expected improvement (EI) when constraints are incorporated. However, the theoretical convergence rate of CEI has not been established. In this work, we study the convergence rate of CEI by analyzing its simple regret upper bound. First, we show that when the objective function $f$ and constraint function $c$ are assumed to each lie in a reproducing kernel Hilbert space (RKHS), CEI achieves the convergence rates of $\mathcal{O} \left(t^{-\frac{1}{2}}\log^{\frac{d+1}{2}}(t) \right) \ \text{and }\ \mathcal{O}\left(t^{\frac{-ν}{2ν+d}} \log^{\fracν{2ν+d}}(t)\right)$ for the commonly used squared exponential and Matérn kernels ($ν>\frac{1}{2}$), respectively. Second, we show that when $f$ is assumed to be sampled from Gaussian processes (GPs), CEI achieves similar convergence rates with a high probability. Numerical experiments are performed to validate the theoretical analysis.

stat.ML↗
arXiv2026

Specific Emitter Identification via Active Learning

With the rapid growth of wireless communications, specific emitter identification (SEI) is significant for communication security. However, its model training relies heavily on the large-scale labeled data, which are costly and time-consuming to obtain. To address this challenge, we propose an SEI approach enhanced by active learning (AL), which follows a three-stage semi-supervised training scheme. In the first stage, self-supervised contrastive learning is employed with a dynamic dictionary update mechanism to extract robust representations from large amounts of the unlabeled data. In the second stage, supervised training on a small labeled dataset is performed, where the contrastive and cross-entropy losses are jointly optimized to improve the feature separability and strengthen the classification boundaries. In the third stage, an AL module selects the most valuable samples from the unlabeled data for annotation based on the uncertainty and representativeness criteria, further enhancing generalization under limited labeling budgets. Experimental results on the ADS-B and WiFi datasets demonstrate that the proposed SEI approach significantly outperforms the conventional supervised and semi-supervised methods under limited annotation conditions, achieving higher recognition accuracy with lower labeling cost.

cs.AI↗
arXiv2026

Secure Communication via Modulation Order Confusion

With the increasing threat posed by modulation classification to wireless security, this paper proposes a secure communication framework based on modulation order confusion (MOC), which intentionally disguises the original modulation as a higher- or lower-order one to mislead eavesdroppers. For single-antenna systems, two schemes are developed: symbol random mapping and symbol time diversity, enabling modulation order confusion with customized receivers. For multi-antenna systems, receiver-transparent MOC schemes are proposed, including series-expansion-based and constellation-path-based signal designs, and are further extended to RIS-assisted systems with joint beamformer and RIS reflection design. Numerical results show that the proposed schemes effectively defeat both deep-learning-based and expert-knowledge-based modulation classifiers without degrading communication performance.

cs.IT↗
arXiv2025

Provable Repair of Deep Neural Network Defects by Preimage Synthesis and Property Refinement

It is known that deep neural networks may exhibit dangerous behaviors under various security threats (e.g., backdoor attacks, adversarial attacks and safety property violation) and there exists an ongoing arms race between attackers and defenders. In this work, we propose a complementary perspective to utilize recent progress on "neural network repair" to mitigate these security threats and repair various kinds of neural network defects (arising from different security threats) within a unified framework, offering a potential silver bullet solution to real-world scenarios. To substantially push the boundary of existing repair techniques (suffering from limitations such as lack of guarantees, limited scalability, considerable overhead, etc) in addressing more practical contexts, we propose ProRepair, a novel provable neural network repair framework driven by formal preimage synthesis and property refinement. The key intuitions are: (i) synthesizing a precise proxy box to characterize the feature space preimage, which can derive a bounded distance term sufficient to guide the subsequent repair step towards the correct outputs, and (ii) performing property refinement to enable surgical corrections and scale to more complex tasks. We evaluate ProRepair across four security threats repair tasks on six benchmarks and the results demonstrate it outperforms existing methods in effectiveness, efficiency and scalability. For point-wise repair, ProRepair corrects models while preserving performance and achieving significantly improved generalization, with a speedup of 5x to 2000x over existing provable approaches. In region-wise repair, ProRepair successfully repairs all 36 safety property violation instances (compared to 8 by the best existing method), and can handle 18x higher dimensional spaces.

cs.CR↗
arXiv2025

Generalized Security-Preserving Refinement for Concurrent Systems

Ensuring compliance with Information Flow Security (IFS) is known to be challenging, especially for concurrent systems with large codebases such as multicore operating system (OS) kernels. Refinement, which verifies that an implementation preserves certain properties of a more abstract specification, is promising for tackling such challenges. However, in terms of refinement-based verification of security properties, existing techniques are still restricted to sequential systems or lack the expressiveness needed to capture complex security policies for concurrent systems. In this work, we present a generalized security-preserving refinement technique, particularly for verifying the IFS of concurrent systems governed by potentially complex security policies. We formalize the IFS properties for concurrent systems and present a refinement-based compositional approach to prove that the generalized security properties (e.g., intransitive noninterference) are preserved between implementation and abstraction. The key intuition enabling such reasoning, compared to previous refinement work, is to establish a step-mapping relation between the implementation and the abstraction, which is sufficient to ensure that every paired step (in the abstraction and the implementation, respectively) is either permitted or prohibited by the security policy. We apply our approach to verify two non-trivial case studies against a collection of security policies. Our proofs are fully mechanized in Isabelle/HOL, during which we identified that two covert channels previously reported in the ARINC 653 single-core standard also exist in the ARINC 653 multicore standard. We subsequently proved the correctness of the revised mechanism, showcasing the effectiveness of our approach.

cs.LO↗