Yan Li

This study estimates the true social cost of corporate data breaches by quantifying the direct financial and opportunity costs to victims, finding that these costs can significantly exceed corporate settlements, though the marginal social cost per victim appears to be declining over time.

This paper conducts the first real-world safety evaluation of the personal AI agent OpenClaw, demonstrating that its broad system access creates inherent vulnerabilities that significantly increase the attack success rate regardless of the underlying large language model.

E-MIA introduces a novel, stealthy black-box membership inference attack that converts verifiable hard evidence within a candidate document into an objective, multi-part exam score to determine if the document was ingested into a RAG knowledge base.

The paper proposes SALO, a novel detector that monitors the dynamic, layer-wise activation pattern (Refusal Trajectory) to improve jailbreak detection robustness compared to traditional methods relying on static terminal representations.

The paper introduces FraudBench, a multimodal benchmark designed to detect AI-generated fraudulent refund evidence, finding that current AI models struggle significantly with claim-conditioned fake-damage detection.

The paper proposes SubPopMark, a novel subpopulation-driven framework that injects harmless, verifiable markers into distilled datasets to prevent copyright infringement and data leakage.

LymphNode is a novel, post-hoc access control framework that protects Deep Neural Networks (DNNs) from model extraction and inversion attacks by enforcing a default-deny policy and selectively restoring utility only for authorized queries.

The paper introduces BYOT-CPS, a hybrid cyber-physical testbed that bridges the gap between purely simulated and purely physical IoT testing environments, enabling realistic and scalable security assessment.

CyBOKClaw is an interpretable human-in-the-loop retrieval framework designed to map broad cybersecurity keywords to the Cyber Security Body of Knowledge (CyBOK), achieving high expert-guided mapping accuracy on both development and validation datasets.

The paper introduces Metacognitive Memory Policy Optimization (MMPO), a novel memory training approach that optimizes LLM memory not based on final task success, but on minimizing epistemic uncertainty in intermediate summaries, significantly improving long-horizon agent performance.

The paper proposes Meta-Team, an experience-driven framework that enables multi-agent systems (MAS) to collaboratively self-evolve by transforming complex execution experiences into reusable improvements for agent behaviors and coordination.

The paper introduces FAM-Bench, a novel multimodal benchmark designed to test advanced, condition-aware reasoning for food-as-medicine applications.

GSAM introduces a generalizable and safe robotic framework for articulated object manipulation, significantly improving success rates and reducing variability across diverse tasks by integrating commonsense reasoning and explicit collision constraints.

CAREAgent is a novel agent designed for fine-grained clinical order generation, achieving significant performance improvements on unseen benchmarks by integrating structured reasoning and tool usage.

MViewRouter proposes a multi-view framework that internalizes geometric equivariance using a Multi-view Alternating Attention mechanism to improve generalization and stabilize training for combinatorial routing problems like TSP and CVRP.

The paper proposes DART, a test-time adaptation method that enhances zero-resource dense retrieval reranking by adaptively tuning a bilinear scoring matrix using pseudo-positive and pseudo-negative examples, achieving significant performance gains with minimal latency.

This survey reviews how Large and Multi-modal Language Models (LLMs/MM-LLMs) are being applied to integrate diverse data sources for enhanced decision support in transportation systems management and operations.

The paper introduces Moment-Video, a new benchmark that diagnoses the ability of video MLLMs to understand brief, critical visual events, revealing that current models struggle significantly with temporal fidelity.

This paper investigates whether adults' struggles with conjunctive causal rules persist when they have agency through active exploration.