Description

本日紹介した論文の一覧

• Invisible Prompts, Visible Threats: Malicious Font Injection in External
  Resources for Large Language Models
  http://arxiv.org/abs/2505.16957v1

• Backdoor Cleaning without External Guidance in MLLM Fine-tuning
  http://arxiv.org/abs/2505.16916v1

• CAIN: Hijacking LLM-Humans Conversations via a Two-Stage Malicious
  System Prompt Generation and Refining Framework
  http://arxiv.org/abs/2505.16888v1

• Unlearning Isn't Deletion: Investigating Reversibility of Machine
  Unlearning in LLMs
  http://arxiv.org/abs/2505.16831v1

• CoTSRF: Utilize Chain of Thought as Stealthy and Robust Fingerprint of
  Large Language Models
  http://arxiv.org/abs/2505.16785v1

• When Safety Detectors Aren't Enough: A Stealthy and Effective Jailbreak
  Attack on LLMs via Steganographic Techniques
  http://arxiv.org/abs/2505.16765v1

• Mitigating Fine-tuning Risks in LLMs via Safety-Aware Probing
  Optimization
  http://arxiv.org/abs/2505.16737v1

• Robust LLM Fingerprinting via Domain-Specific Watermarks
  http://arxiv.org/abs/2505.16723v1

• BitHydra: Towards Bit-flip Inference Cost Attack against Large Language
  Models
  http://arxiv.org/abs/2505.16670v1

• Unsupervised Network Anomaly Detection with Autoencoders and Traffic
  Images
  http://arxiv.org/abs/2505.16650v1

• BadVLA: Towards Backdoor Attacks on Vision-Language-Action Models via
  Objective-Decoupled Optimization
  http://arxiv.org/abs/2505.16640v1

• Energy Consumption Framework and Analysis of Post-Quantum Key-Generation
  on Embedded Devices
  http://arxiv.org/abs/2505.16614v1

• Finetuning-Activated Backdoors in LLMs
  http://arxiv.org/abs/2505.16567v1

• CTRAP: Embedding Collapse Trap to Safeguard Large Language Models from
  Harmful Fine-Tuning
  http://arxiv.org/abs/2505.16559v1

• DuFFin: A Dual-Level Fingerprinting Framework for LLMs IP Protection
  http://arxiv.org/abs/2505.16530v1

• Language-based Security and Time-inserting Supervisor
  http://arxiv.org/abs/2505.16503v1

• Password Strength Detection via Machine Learning: Analysis, Modeling,
  and Evaluation
  http://arxiv.org/abs/2505.16439v1

• Consistent and Compatible Modelling of Cyber Intrusions and Incident
  Response Demonstrated in the Context of Malware Attacks on Critical
  Infrastructure
  http://arxiv.org/abs/2505.16398v1

• Privacy-Aware Cyberterrorism Network Analysis using Graph Neural
  Networks and Federated Learning
  http://arxiv.org/abs/2505.16371v1

• ReCopilot: Reverse Engineering Copilot in Binary Analysis
  http://arxiv.org/abs/2505.16366v1

• SuperPure: Efficient Purification of Localized and Distributed
  Adversarial Patches via Super-Resolution GAN Models
  http://arxiv.org/abs/2505.16318v1

• Poster: Towards an Automated Security Testing Framework for Industrial
  UEs
  http://arxiv.org/abs/2505.16300v1

• All You Need is "Leet": Evading Hate-speech Detection AI
  http://arxiv.org/abs/2505.16263v1

• Interpretable Anomaly Detection in Encrypted Traffic Using SHAP with
  Machine Learning Models
  http://arxiv.org/abs/2505.16261v1

• Verifying Differentially Private Median Estimation
  http://arxiv.org/abs/2505.16246v1

• A Scalable Hierarchical Intrusion Detection System for Internet of
  Vehicles
  http://arxiv.org/abs/2505.16215v1

• VIVID: A Novel Approach to Remediation Prioritization in Static
  Application Security Testing (SAST)
  http://arxiv.org/abs/2505.16205v1

• SafeKey: Amplifying Aha-Moment Insights for Safety Reasoning
  http://arxiv.org/abs/2505.16186v1

• Outsourcing SAT-based Verification Computations in Network Security
  http://arxiv.org/abs/2505.16137v1

• Exte