arXiv subjects

Xuran Li

Publications and source records attributed to Xuran Li.

10 recordsLinked to original sources

arXiv2026

Evaluating LLM Personalization via Semantic Constraint Verification

Current evaluation paradigms for Large Language Model (LLM) personalization rely heavily on brittle surface-matching metrics or computationally expensive LLM-as-a-judge protocols, both of which lack interpretability. To address these limitations, we introduce Natural Language Inference Constraint Verification (NLICV), a scalable, semantically invariant framework that maps sentence meanings to truth-condition sets to verify personalization constraints via a Natural Language Inference (NLI) model. Moving beyond binary scoring, NLICV categorizes LLM behaviors into four distinct modes: personalization, generalization, sycophancy, and failure. Extensive experiments demonstrate that NLICV aligns closely with human annotations while drastically reducing the latency and token costs associated with LLM judges (up to 2100 inference speedup). Finally, through an ablation-based procedure, NLICV pinpoints the exact sentences driving the constraint verification, yielding faithful, understandable evidence for its evaluations.

cs.CL↗

arXiv2025

An Empirical Study of Sample Selection Strategies for Large Language Model Repair

Large language models (LLMs) are increasingly deployed in real-world systems, yet they can produce toxic or biased outputs that undermine safety and trust. Post-hoc model repair provides a practical remedy, but the high cost of parameter updates motivates selective use of repair data. Despite extensive prior work on data selection for model training, it remains unclear which sampling criteria are most effective and efficient when applied specifically to behavioral repair of large generative models. Our study presents a systematic analysis of sample prioritization strategies for LLM repair. We evaluate five representative selection methods, including random sampling, K-Center, gradient-norm-based selection(GraNd), stratified coverage (CCS), and a Semantic-Aware Prioritized Sampling (SAPS) approach we proposed. Repair effectiveness and trade-offs are assessed through toxicity reduction, perplexity on WikiText-2 and LAMBADA, and three composite metrics: the Repair Proximity Score (RPS), the Overall Performance Score (OPS), and the Repair Efficiency Score (RES). Experimental results show that SAPS achieves the best balance between detoxification, utility preservation, and efficiency, delivering comparable or superior repair outcomes with substantially less data. Random sampling remains effective for large or robust models, while high-overhead methods such as CCS and GraNd provide limited benefit. The optimal data proportion depends on model scale and repair method, indicating that sample selection should be regarded as a tunable component of repair pipelines. Overall, these findings establish selection-based repair as an efficient and scalable paradigm for maintaining LLM reliability.

cs.LG↗

arXiv2025

PRUNE: A Patching Based Repair Framework for Certifiable Unlearning of Neural Networks

It is often desirable to remove (a.k.a. unlearn) a specific part of the training data from a trained neural network model. A typical application scenario is to protect the data holder's right to be forgotten, which has been promoted by many recent regulation rules. Existing unlearning methods involve training alternative models with remaining data, which may be costly and challenging to verify from the data holder or a thirdparty auditor's perspective. In this work, we provide a new angle and propose a novel unlearning approach by imposing carefully crafted "patch" on the original neural network to achieve targeted "forgetting" of the requested data to delete. Specifically, inspired by the research line of neural network repair, we propose to strategically seek a lightweight minimum "patch" for unlearning a given data point with certifiable guarantee. Furthermore, to unlearn a considerable amount of data points (or an entire class), we propose to iteratively select a small subset of representative data points to unlearn, which achieves the effect of unlearning the whole set. Extensive experiments on multiple categorical datasets demonstrates our approach's effectiveness, achieving measurable unlearning while preserving the model's performance and being competitive in efficiency and memory consumption compared to various baseline methods.

cs.LG↗

arXiv2024

Perturbation Effects on Robustness and Individual Fairness

Deep neural networks are vulnerable to adversarial perturbations that can simultaneously degrade prediction robustness and individual fairness across diverse application settings. However, existing evaluation protocols typically assess these dimensions in isolation, thereby obscuring critical failure modes. To bridge this gap, we formalize Robust Individual Fairness (RIF): under semantic-preserving (truth-condition-preserving) perturbations, predictions should remain both correct with respect to the ground truth and invariant across semantically equivalent individuals. To surface RIF violations in practice, we introduce RIFair, a black-box adversarial framework that leverages a decoupled perturbation strategy to construct semantically preserved yet unrobust and/or unfair instance pairs. Experiments across multiple model architectures and real-world textual datasets show that robustness-only or fairness-only metrics often miss Robust Biased and Unrobust Fair behaviors. RIFair}reliably exposes these hidden vulnerabilities, supporting RIF as a necessary criterion for trustworthy model assessment. The experimental code is publicly available at https://github.com/Xuran-LI/RIFair.

cs.LG↗

arXiv2023

RobustFair: Adversarial Evaluation through Fairness Confusion Directed Gradient Search

Deep neural networks (DNNs) often face challenges due to their vulnerability to various adversarial perturbations, including false perturbations that undermine prediction accuracy and biased perturbations that cause biased predictions for similar inputs. This paper introduces a novel approach, RobustFair, to evaluate the accurate fairness of DNNs when subjected to these false or biased perturbations. RobustFair employs the notion of the fairness confusion matrix induced in accurate fairness to identify the crucial input features for perturbations. This matrix categorizes predictions as true fair, true biased, false fair, and false biased, and the perturbations guided by it can produce a dual impact on instances and their similar counterparts to either undermine prediction accuracy (robustness) or cause biased predictions (individual fairness). RobustFair then infers the ground truth of these generated adversarial instances based on their loss function values approximated by the total derivative. To leverage the generated instances for trustworthiness improvement, RobustFair further proposes a data augmentation strategy to prioritize adversarial instances resembling the original training set, for data augmentation and model retraining. Notably, RobustFair excels at detecting intertwined issues of robustness and individual fairness, which are frequently overlooked in standard robustness and individual fairness evaluations. This capability empowers RobustFair to enhance both robustness and individual fairness evaluations by concurrently identifying defects in either domain. Empirical case studies and quantile regression analyses on benchmark datasets demonstrate the effectiveness of the fairness confusion matrix guided perturbation for false or biased adversarial instance generation.

cs.LG↗

arXiv2023

On the Performance Trade-off of Distributed Integrated Sensing and Communication Networks

In this letter, we analyze the performance trade-off in distributed integrated sensing and communication (ISAC) networks. Specifically, with the aid of stochastic geometry theory, we derive the probability of detection of that of the coverage given user number. Based on the analytical derivations, we provide a quantitative description of the performance limits and the performance trade-off between sensing and communication in a distributed ISAC network under the given transmit power and bandwidth budget. Extensive simulations are conducted and the numerical results validate the accuracy of our derivations.

cs.IT↗

arXiv2023

Privacy and Security in Ubiquitous Integrated Sensing and Communication: Threats, Challenges and Future Directions

Integrated sensing and communication (ISAC) technology is one of the featuring technologies of the next-generation communication systems. When sensing capability becomes ubiquitous, more information can be collected, which can facilitate many applications in intelligent transportation, unmanned aerial vehicle (UAV) surveillance and healthcare. However, it also faces many information privacy leakage and security issues. This article highlights the potential threats to privacy and security and the technical challenges to realizing private and secure ISAC. Three promising combating solutions including artificial intelligence (AI)-enabled schemes, friendly jamming and reconfigurable intelligent surface (RIS)-assisted design are provided to maintain user privacy and ensure information security. Case studies demonstrate their effectiveness.

eess.SP↗

arXiv2022

Accurate Fairness: Improving Individual Fairness without Trading Accuracy

Accuracy and individual fairness are both crucial for trustworthy machine learning, but these two aspects are often incompatible with each other so that enhancing one aspect may sacrifice the other inevitably with side effects of true bias or false fairness. We propose in this paper a new fairness criterion, accurate fairness, to align individual fairness with accuracy. Informally, it requires the treatments of an individual and the individual's similar counterparts to conform to a uniform target, i.e., the ground truth of the individual. We prove that accurate fairness also implies typical group fairness criteria over a union of similar sub-populations. We then present a Siamese fairness in-processing approach to minimize the accuracy and fairness losses of a machine learning model under the accurate fairness constraints. To the best of our knowledge, this is the first time that a Siamese approach is adapted for bias mitigation. We also propose fairness confusion matrix-based metrics, fair-precision, fair-recall, and fair-F1 score, to quantify a trade-off between accuracy and individual fairness. Comparative case studies with popular fairness datasets show that our Siamese fairness approach can achieve on average 1.02%-8.78% higher individual fairness (in terms of fairness through awareness) and 8.38%-13.69% higher accuracy, as well as 10.09%-20.57% higher true fair rate, and 5.43%-10.01% higher fair-F1 score, than the state-of-the-art bias mitigation techniques. This demonstrates that our Siamese fairness approach can indeed improve individual fairness without trading accuracy. Finally, the accurate fairness criterion and Siamese fairness approach are applied to mitigate the possible service discrimination with a real Ctrip dataset, by on average fairly serving 112.33% more customers (specifically, 81.29% more customers in an accurately fair way) than baseline models.

cs.LG↗

arXiv2022

Infectious Probability Analysis on COVID-19 Spreading with Wireless Edge Networks

The emergence of infectious disease COVID-19 has challenged and changed the world in an unprecedented manner. The integration of wireless networks with edge computing (namely wireless edge networks) brings opportunities to address this crisis. In this paper, we aim to investigate the prediction of the infectious probability and propose precautionary measures against COVID-19 with the assistance of wireless edge networks. Due to the availability of the recorded detention time and the density of individuals within a wireless edge network, we propose a stochastic geometry-based method to analyze the infectious probability of individuals. The proposed method can well keep the privacy of individuals in the system since it does not require to know the location or trajectory of each individual. Moreover, we also consider three types of mobility models and the static model of individuals. Numerical results show that analytical results well match with simulation results, thereby validating the accuracy of the proposed model. Moreover, numerical results also offer many insightful implications. Thereafter, we also offer a number of countermeasures against the spread of COVID-19 based on wireless edge networks. This study lays the foundation toward predicting the infectious risk in realistic environment and points out directions in mitigating the spread of infectious diseases with the aid of wireless edge networks.

cs.SI↗

arXiv2022

Is Blockchain for Internet of Medical Things a Panacea for COVID-19 Pandemic?

The outbreak of the COVID-19 pandemic has deeply influenced the lifestyle of the general public and the healthcare system of the society. As a promising approach to address the emerging challenges caused by the epidemic of infectious diseases like COVID-19, Internet of Medical Things (IoMT) deployed in hospitals, clinics, and healthcare centers can save the diagnosis time and improve the efficiency of medical resources though privacy and security concerns of IoMT stall the wide adoption. In order to tackle the privacy, security, and interoperability issues of IoMT, we propose a framework of blockchain-enabled IoMT by introducing blockchain to incumbent IoMT systems. In this paper, we review the benefits of this architecture and illustrate the opportunities brought by blockchain-enabled IoMT. We also provide use cases of blockchain-enabled IoMT on fighting against the COVID-19 pandemic, including the prevention of infectious diseases, location sharing and contact tracing, and the supply chain of injectable medicines. We also outline future work in this area.

cs.DC↗