SCARA is a novel, end-to-end framework that autonomously connects binary-level vulnerability candidates to conditionally validated remedies for opaque industrial software, achieving high precision and…

The paper introduces APIOT, the first LLM framework capable of autonomously performing the full discovery, exploitation, patching, and verification cycle against bare-metal industrial OT devices.

The paper argues that post-hoc mitigation techniques like machine unlearning are insufficient to cure legal liability arising from the unlawful acquisition and training on copyrighted data, advocating…

The paper introduces retraining-free frameworks (Meow2X and TRNE) that mechanistically localize and suppress toxicity within language models by analyzing activation differences, achieving safety impro…

RefineRAG introduces a novel word-level poisoning framework that significantly enhances knowledge poisoning attacks against RAG systems, achieving state-of-the-art effectiveness and transferability to…

The paper introduces SONAR, a prompt sanitization framework that uses natural language inference metrics to identify and remove malicious instructions injected into LLM prompts, achieving near-zero at…

The paper introduces ToxiAlert-Bench, a large-scale audio dataset that uniquely annotates both textual and paralinguistic sources of toxicity, and proposes a dual-head neural network that significantl…

SHIELDS is a multi-agent system that uses LLMs to automate OS hardening by iteratively proposing and refining fixes based on real-time system feedback, achieving up to 73% remediation success.

The paper reframes manufacturing ransomware recovery from a simple backup restoration task to a complex critical-infrastructure continuity problem, proposing Minimum Viable Factory Recovery (MVF Recov…

This paper identifies and analyzes unintentional cross-user contamination (UCC), a failure mode where benign, scope-bound artifacts degrade the outcomes of different users in shared-state LLM agents,…

The paper argues that current LLM benchmark datasets are often contaminated by being included in pretraining data, and proposes that future benchmarks must be contamination-resistant and support infer…

SCRIBE is a novel framework that enables reliable source-level patching of binaries by performing 'binary-aware' recompilation, successfully resolving syntactic and semantic inaccuracies inherent in d…

The paper introduces 'contrastive privacy,' a formal, model-agnostic, and quantitative method for evaluating the semantic success of AI-based sanitization across multiple media modalities.

The paper introduces Decaf, a system that uses automatic feedback and search to significantly improve the semantic correctness and accuracy of neural decompilers, boosting the decompilation rate from…

The paper introduces REBench, a comprehensive, standardized benchmark dataset designed to enable fair and rigorous evaluation of Large Language Models (LLMs) on complex binary reverse engineering task…

肖代替了视觉令牌的永久删除，通过可恢复的路由来改进视觉语言模型的性能

RASER introduces a family of cheap, router-based systems that selectively decide whether to perform expensive multi-hop retrieval, significantly reducing LLM token costs while maintaining state-of-the…

This review surveys advanced techniques—including generative models, multimodal learning, and closed-loop workflows—for automated inverse materials design, enabling the targeted discovery of novel cry…

This perspective reviews advanced design principles for Metal-Organic Frameworks (MOFs) used in water harvesting and details how integrating Artificial Intelligence (AI) can accelerate the discovery o…

The paper proposes LaRA, a layer-wise representation analysis framework that detects data contamination in RL post-trained LLMs by analyzing geometric deviations across model layers.