arXiv subjects

Ling Shi

Publications and source records attributed to Ling Shi.

At least 37 records · Page 2Linked to original sources

arXiv2026

Variational Robust Kalman Filters: A Unified Framework

Robustness and adaptivity are two competing objectives in Kalman filters (KF). Robustness involves temporarily inflating prior estimates of noise covariances, while adaptivity updates prior beliefs by exploiting measurements. In practical applications, both process and measurement noise can be influenced by outliers, be time-varying, or both. In this work, we propose a variational robust Kalman filter, built on a Student's $t$-distribution induced loss function and variational inference, and solved in a computationally efficient manner. We demonstrate that robustness can be understood as a prerequisite for adaptivity, making it possible to merge the above two competing goals into a single framework through a probabilistic switching rule. Additionally, our proposed filter can recover conventional KF, robust KF, and adaptive KF by tuning parameters, and can suppress both the imperfect process and measurement noise, enabling it to perform superiorly in complex noise environments. Simulations verify the effectiveness of the proposed method.

cs.IT↗

arXiv2026

SixthSense: Task-Agnostic Proprioception-Only Whole-Body Wrench Estimation for Humanoids

Humanoid robots are entering our physical world at scale, yet as oversized toys--good at singing and dancing, but short on force-interaction capabilities for practical tasks. Bridging this gap necessitates prioritizing reliable contact perception as a fundamental requirement. Estimating external wrenches in humanoids is complicated by floating-base dynamics and indeterminate contact locations. Existing analytical frameworks require idealistic assumptions and hard-to-obtain measurements, which are often unavailable in practice. To bridge this gap, we propose SixthSense, a task-agnostic approach that infers whole-body contact timing, location, and wrenches from proprioception and IMU data alone. To capture the multi-modal dynamics between unstructured contact inputs and the uncertain motion outputs, we employ conditional flow matching to tokenize proprioceptive histories and estimate a spatiotemporally sparse contact-event flow. SixthSense serves as a plug-and-play perception module for applications including collision detection, physical human-robot interaction, and force-feedback teleoperation. Experiments across standing, walking, and whole-body motion-tracking policies showcased unprecedented performance in diverse behaviors.

cs.RO↗

arXiv2026

From Insight to Action: A Novel Framework for Interpretability-Guided Data Selection in Large Language Models

While mechanistic interpretability tools like Sparse Autoencoders (SAEs) can uncover meaningful features within Large Language Models (LLMs), a critical gap remains in transforming these insights into practical actions for model optimization. We bridge this gap with the hypothesis that data selection guided by a model's internal task features is a effective training strategy. Inspired by this, we propose Interpretability-Guided Data Selection (IGDS), a framework that first identifies these causal task features through frequency recall and interventional filtering, then selects ``Feature-Resonant Data'' that maximally activates task features for fine-tuning. We validate IGDS on mathematical reasoning, summarization, and translation tasks within Gemma-2, LLaMA-3.1, and Qwen3 models. Our experiments demonstrate exceptional data efficiency: on the Math task, IGDS surpasses full-dataset fine-tuning by a remarkable 17.4% on Gemma-2-2B while using only 50% of the data, and outperforms established baselines focused on data quality and diversity. Analysis confirms a strong positive correlation between feature amplification and task performance improvement. IGDS thus provides a direct and effective framework to enhance LLMs by leveraging their internal mechanisms, validating our core hypothesis.

cs.AI↗

arXiv2026

Uncovering Logit Suppression Vulnerabilities in LLM Safety Alignment

Large language models (LLMs) have revolutionized various applications, making robust safety alignment essential to prevent harmful outputs. Current safety alignment techniques, however, harbor inherent vulnerabilities due to their reliance on logit suppression. In this work, we identify critical logit-level vulnerabilities by introducing Semantic-sensitive Alignment and Generation (SSAG), a method designed to systematically manipulate output-layer logits without altering model parameters. Experiments on five popular LLMs show that SSAG exposes harmful responses with a 95% success rate while reducing response time by 86%. VulMine also demonstrates superior attack efficacy, achieving an average ASR of up to 77% against strong defensive mechanisms. These findings reveal crucial weaknesses in existing alignment methods, highlighting an urgent need for improved vulnerability detection and robust safety alignment strategies. Our code is available on github.

cs.CR↗

arXiv2026

SAGE: A Service Agent Graph-guided Evaluation Benchmark

The development of Large Language Models (LLMs) has catalyzed automation in customer service, yet benchmarking their performance remains challenging. Existing benchmarks predominantly rely on static paradigms and single-dimensional metrics, failing to account for diverse user behaviors or the strict adherence to structured Standard Operating Procedures (SOPs) required in real-world deployments. To bridge this gap, we propose SAGE (Service Agent Graph-guided Evaluation), a universal multi-agent benchmark for automated, dual-axis assessment. SAGE formalizes unstructured SOPs into Dynamic Dialogue Graphs, enabling precise verification of logical compliance and comprehensive path coverage. We introduce an Adversarial Intent Taxonomy and a modular Extension Mechanism, enabling low-cost deployment across domains and facilitating automated dialogue data synthesis. Evaluation is conducted via a framework where Judge Agents and a Rule Engine analyze interactions between User and Service Agents to generate deterministic ground truth. Extensive experiments on 27 LLMs across 6 industrial scenarios reveal a significant ``Execution Gap'' where models accurately classify intents but fail to derive correct subsequent actions. We also observe ``Empathy Resilience'', a phenomenon where models maintain polite conversational facades despite underlying logical failures under high adversarial intensity. Code and resources are available at https://anonymous.4open.science/r/SAGE-Bench-4CD3/.

cs.AI↗

arXiv2026

Exposing the Ghost in the Transformer: Abnormal Detection for Large Language Models via Hidden State Forensics

The widespread adoption of Large Language Models (LLMs) in critical applications has introduced severe reliability and security risks, as LLMs remain vulnerable to notorious threats such as hallucinations, jailbreak attacks, and backdoor exploits. These vulnerabilities have been weaponized by malicious actors, leading to unauthorized access, widespread misinformation, and compromised LLM-embedded system integrity. In this work, we introduce a novel approach to detecting abnormal behaviors in LLMs via hidden state forensics. By systematically inspecting layer-specific activation patterns, we develop a general framework that can efficiently identify a range of security threats in real-time without imposing prohibitive computational costs. Extensive experiments indicate detection accuracies exceeding 95% and consistently robust performance across multiple models in most scenarios, while preserving the ability to detect novel attacks effectively. Furthermore, the computational overhead remains minimal, with detector inference taking merely fractions of a second. The significance of this work lies in proposing a promising strategy to reinforce the security of LLM-integrated systems, paving the way for safer and more reliable deployment in high-stakes domains. By enabling real-time detection that can also support the mitigation of abnormal behaviors, it represents a meaningful step toward ensuring the trustworthiness of AI systems amid rising security challenges.

cs.CR↗

arXiv2026

Distributed State Estimation for Discrete-Time Linear Systems over Directed Graphs: A Measurement Perspective

This paper proposes a novel consensus-based distributed filter over directed graphs under the collectively observability condition. The distributed filter is designed using an augmented leader-following information fusion strategy, and the gain parameter is determined exclusively using local information. Additionally, the lower bound of the fusion step number is derived to ensure that the estimation error covariance remains uniformly upper-bounded. Furthermore, the lower bounds for the convergence rates of the steady-state performance gap between the proposed filter and the centralized filter are provided as the fusion step number approaches infinity. The analysis demonstrates that the convergence rate is at least as fast as exponential convergence, provided the communication topology satisfies the spectral norm condition. Finally, the theoretical results are validated through two simulation examples.

eess.SY↗

arXiv2026

Data-Driven Structured Controller Design Using the Matrix S-Procedure

This paper focuses on the data-driven optimal structured controller design for discrete-time linear time-invariant (LTI) systems, considering both the $H_2$ performance and the $H_\infty$ performance. Specifically, we consider three scenarios: (i) the model-based structured control, (ii) the data-driven unstructured control, and (iii) the data-driven structured control. For the $H_2$ performance, we primarily investigate cases (ii) and (iii), since case (i) has been extensively studied in the literature. For the $H_\infty$ performance, all three scenarios are considered. For the structured control, we introduce a linearization technique that transforms the original nonconvex problem into a semidefinite programming (SDP) problem. Based on this transformation, we develop an iterative linear matrix inequality (ILMI) algorithm. For the data-driven control, we describe the set of all possible system matrices that can generate the sequence of collected data. Additionally, we propose a sufficient condition to handle all possible system matrices using the matrix S-procedure. The data-driven structured control is followed by combining the previous two cases. We compare our methods with those in the existing literature and demonstrate our superiority via several numerical simulations.

math.OC↗

arXiv2026

DEP: A Decentralized Large Language Model Evaluation Protocol

With the rapid development of Large Language Models (LLMs), a large number of benchmarks have been proposed. However, most benchmarks lack unified evaluation standard and require the manual implementation of custom scripts, making results hard to ensure consistency and reproducibility. Furthermore, mainstream evaluation frameworks are centralized, with datasets and answers, which increases the risk of benchmark leakage. To address these issues, we propose a Decentralized Evaluation Protocol (DEP), a decentralized yet unified and standardized evaluation framework through a matching server without constraining benchmarks. The server can be mounted locally or deployed remotely, and once adapted, it can be reused over the long term. By decoupling users, LLMs, and benchmarks, DEP enables modular, plug-and-play evaluation: benchmark files and evaluation logic stay exclusively on the server side. In remote setting, users cannot access the ground truth, thereby achieving data isolation and leak-proof evaluation. To facilitate practical adoption, we develop DEP Toolkit, a protocol-compatible toolkit that supports features such as breakpoint resume, concurrent requests, and congestion control. We also provide detailed documentation for adapting new benchmarks to DEP. Using DEP toolkit, we evaluate multiple LLMs across benchmarks. Experimental results verify the effectiveness of DEP and show that it reduces the cost of deploying benchmark evaluations. As of February 2026, we have adapted over 60 benchmarks and continue to promote community co-construction to support unified evaluation across various tasks and domains.

cs.CL↗

arXiv2026

Reinforcing Real-world Service Agents: Balancing Utility and Cost in Task-oriented Dialogue

The rapid evolution of Large Language Models (LLMs) has accelerated the transition from conversational chatbots to general agents. However, effectively balancing empathetic communication with budget-aware decision-making remains an open challenge. Since existing methods fail to capture these complex strategic trade-offs, we propose InteractCS-RL, a framework that reframes task-oriented dialogue as a multi-granularity reinforcement learning process. Specifically, we first establish a User-centric Interaction Framework to provide a high-fidelity training gym, enabling agents to dynamically explore diverse strategies with persona-driven users. Then, we introduce Cost-aware Multi-turn Policy Optimization (CMPO) with a hybrid advantage estimation strategy. By integrating generative process credits and employing a PID-Lagrangian cost controller, CMPO effectively guides the policy to explore Pareto boundary between user reward and global cost constraints. Extensive experiments on customized real business scenarios demonstrate that InteractCS-RL significantly outperform other baselines across three evaluation dimensions. Further evaluation on tool-agent-user interaction benchmarks verify InteractCS-RL robustness across diverse domains.

cs.CL↗

arXiv2026

Online Coreset Selection for Learning Dynamic Systems

With the increasing availability of streaming data in dynamic systems, a critical challenge in data-driven modeling for control is how to efficiently select informative data to characterize system dynamics. In this work, we develop an online coreset selection method for set-membership identification in the presence of process disturbances, improving data efficiency while preserving convergence guarantees. Specifically, we derive a stacked polyhedral representation that over-approximates the feasible parameter set. Based on this representation, we propose a geometric selection criterion that retains a data point only if it induces a sufficient contraction of the feasible set. Theoretically, the feasible-set volume is shown to converge to zero almost surely under persistently exciting data and a tight disturbance bound. When the disturbance bound is mismatched, an explicit Hausdorff-distance bound is derived to quantify the resulting identification error. In addition, an upper bound on the expected coreset size is established and extensions to nonlinear systems with linear-in-the-parameter structures and to bounded measurement noise are discussed. The effectiveness of the proposed method is demonstrated through comprehensive simulation studies.

eess.SY↗

arXiv2026

Data-Driven Structured Control for Continuous-Time LTI Systems

This paper addresses the data-driven structured controller design problem for continuous-time linear time-invariant (LTI) systems. We consider three control objectives, including stabilization, $H_2$ performance, and $H_\infty$ performance. Using the collected data, we construct a minimal matrix ellipsoid that contains all admissible system matrices. We propose some linearization techniques that enable us to incorporate the structural constraint on the controller, which motivates an iterative algorithm for each control objective. Finally, we provide some numerical examples to demonstrate the effectiveness of the proposed methods.

math.OC↗

arXiv2026

Bias-VarianceTrade-off in Kalman Filter-Based Disturbance Observers

The performance of disturbance observers is strongly influenced by the level of prior knowledge about the disturbance model. The simultaneous input and state estimation (SISE) algorithm is widely recognized for providing unbiased minimum-variance estimates under arbitrary disturbance models. In contrast, the Kalman filter-based disturbance observer (KF-DOB) achieves minimum mean-square error estimation when the disturbance model is fully specified. However, practical scenarios often fall between these extremes, where only partial knowledge of the disturbance model is available. This paper investigates the inherent bias-variance trade-off in KF-DOB when the disturbance model is incomplete. We further show that SISE can be interpreted as a special case of KF-DOB, where the disturbance noise covariance tends to infinity. To address this trade-off, we propose two novel estimators: the multi-kernel correntropy Kalman filter-based disturbance observer (MKCKF-DOB) and the interacting multiple models Kalman filter-based disturbance observer (IMMKF-DOB). Simulations verify the effectiveness of the proposed methods.

eess.SP↗

arXiv2026

STEAMROLLER: A Multi-Agent System for Inclusive Automatic Speech Recognition for People who Stutter

People who stutter (PWS) face systemic exclusion in today's voice-driven society, where access to voice assistants, authentication systems, and remote work tools increasingly depends on fluent speech. Current automatic speech recognition (ASR) systems, trained predominantly on fluent speech, fail to serve millions of PWS worldwide. We present STEAMROLLER, a real time system that transforms stuttered speech into fluent output through a novel multi-stage, multi-agent AI pipeline. Our approach addresses three critical technical challenges: (1) the difficulty of direct speech to speech conversion for disfluent input, (2) semantic distortions introduced during ASR transcription of stuttered speech, and (3) latency constraints for real time communication. STEAMROLLER employs a three stage architecture comprising ASR transcription, multi-agent text repair, and speech synthesis, where our core innovation lies in a collaborative multi-agent framework that iteratively refines transcripts while preserving semantic intent. Experiments on the FluencyBank dataset and a user study demonstrates clear word error rate (WER) reduction and strong user satisfaction. Beyond immediate accessibility benefits, fine tuning ASR on STEAMROLLER repaired speech further yields additional WER improvements, creating a pathway toward inclusive AI ecosystems.

cs.CY↗

arXiv2025

Beyond Correctness: Exposing LLM-generated Logical Flaws in Reasoning via Multi-step Automated Theorem Proving

Large Language Models (LLMs) have demonstrated impressive reasoning capabilities, leading to their adoption in high-stakes domains such as healthcare, law, and scientific research. However, their reasoning often contains subtle logical errors masked by fluent language, posing significant risks for critical applications. While existing approaches like fact-checking, self-consistency methods, and rule-based validation provide partial solutions, they fail to detect complex logical flaws in multi-step reasoning. To overcome these challenges, we present MATP, an evaluation framework for systematically verifying LLM reasoning via Multi-step Automatic Theorem Proving. MATP translates natural language reasoning into First-Order Logic (FOL) and applies automated theorem provers to assess step-by-step logical validity. This approach identifies hidden logical errors and provides fine-grained classifications of reasoning correctness. Evaluations on a benchmark comprising 10,830 reasoning instances generated by 10 LLMs across tasks from PrOntoQA-OOD, ProofWriter, and FOLIO show that MATP surpasses prompting-based baselines by over 42 percentage points in reasoning step verification. It further reveals model-level disparities, with reasoning models generating more logically coherent outputs than general models. These results demonstrate MATP's potential to enhance the trustworthiness of LLM-generated reasoning.

cs.SE↗

arXiv2025

Boosting Pointer Analysis With LLM-Enhanced Allocation Function Detection

Pointer analysis is foundational for many static analysis tasks, yet its effectiveness is often hindered by imprecise modeling of heap allocations, particularly in C/C++ programs where custom allocation functions (CAFs) are pervasive. Existing approaches largely overlook these custom allocators, leading to coarse aliasing and low analysis precision. In this paper, we present CAFD, a novel and lightweight technique that enhances pointer analysis by automatically detecting side-effect-free custom allocation functions. CAFD employs a hybrid approach: it uses value-flow analysis to detect straightforward wrappers and leverages Large Language Models (LLMs) to reason about more complex allocation patterns with side effects, ensuring that only side-effect-free functions are modeled as allocators. This targeted enhancement enables precise modeling of heap objects at each call site, achieving context-sensitivity-like benefits without significant overhead. We evaluated CAFD on 17 real-world C projects, identifying over 700 CAFs. Integrating CAFD into a baseline pointer analysis yields a 38x increase in modeled heap objects and a 41.5% reduction in alias set sizes, with only 1.4x runtime overhead. Furthermore, the LLM-enhanced pointer analysis improves indirect call resolution and discovers 29 previously undetected memory bugs, including 6 from real-world industrial applications. These results demonstrate that precise modeling of CAFs has the capability to offer a scalable and practical path to improve pointer analysis in large software systems.

cs.SE↗

arXiv2025

Event-triggered Dual Gradient Tracking for Distributed Resource Allocation

High communication costs create a major bottleneck for distributed resource allocation over unbalanced directed networks. Conventional dual gradient tracking methods, while effective for problems on unbalanced digraphs, rely on periodic communication that creates significant overhead in resource-constrained networks. This paper introduces a novel event-triggered dual gradient tracking algorithm to mitigate this limitation, wherein agents communicate only when local state deviations surpass a predefined threshold. We establish comprehensive convergence guarantees for this approach. First, we prove sublinear convergence for non-convex dual objectives and linear convergence under the Polyak-Łojasiewicz condition. Building on this, we demonstrate that the proposed algorithm achieves sublinear convergence for general strongly convex cost functions and linear convergence for those that are also Lipschitz-smooth. Numerical experiments confirm that our event-triggered method significantly reduces communication events compared to periodic schemes while preserving comparable convergence performance.

eess.SY↗

arXiv2025

High Performance Distributed Control for Large-Scale Linear Systems: A Cover-Based Distributed Observer Approach

In recent years, the distributed-observer-based distributed control law has shown powerful ability to arbitrarily approximate the centralized control performance. However, the traditional distributed observer requires each local observer to reconstruct the state information of the whole system, which is unrealistic for large-scale scenarios. To fill this gap, This paper presents a coverage solution algorithm for large-scale systems that accounts for both physical and communication network characteristics, which can significantly reduce the dimension of local observers. Then, the cover-based distributed observer for large-scale systems is proposed to overcome the problem that the system dynamics are difficult to estimate due to the coupling between cover sets. Furthermore, the two-layer Lyapunov analysis method is adopted and the dynamic transformation lemma of compact errors is proved, which solves the problem of analyzing stability of the error dynamic of the cover-based distributed observer. Finally, it is proved that the distributed control law based on the cover-based distributed observer can also arbitrarily approximate the control performance of the centralized control law, and the dimension of the local observer is greatly reduced compared with the traditional method. The simulation results show the validity of the developed theories.

eess.SY↗