Description

本日紹介した論文の一覧

• TZ-LLM: Protecting On-Device Large Language Models with Arm TrustZone
  https://arxiv.org/abs/2511.13717v1

• Tuning for Two Adversaries: Enhancing the Robustness Against Transfer and Query-Based Attacks using Hyperparameter Tuning
  https://arxiv.org/abs/2511.13654v1

• It's a Feature, Not a Bug: Secure and Auditable State Rollback for Confidential Cloud Applications
  https://arxiv.org/abs/2511.13641v1

• Robust Client-Server Watermarking for Split Federated Learning
  https://arxiv.org/abs/2511.13598v1

• Exploring the Effectiveness of Google Play Store's Privacy Transparency Channels
  https://arxiv.org/abs/2511.13576v1

• ForgeDAN: An Evolutionary Framework for Jailbreaking Aligned Large Language Models
  https://arxiv.org/abs/2511.13548v1

• Interpretable Ransomware Detection Using Hybrid Large Language Models: A Comparative Analysis of BERT, RoBERTa, and DeBERTa Through LIME and SHAP
  https://arxiv.org/abs/2511.13517v1

• Tight and Practical Privacy Auditing for Differentially Private In-Context Learning
  https://arxiv.org/abs/2511.13502v1

• InfoDecom: Decomposing Information for Defending against Privacy Leakage in Split Inference
  https://arxiv.org/abs/2511.13365v1

• Enhancing All-to-X Backdoor Attacks with Optimized Target Class Mapping
  https://arxiv.org/abs/2511.13356v1

• AutoMalDesc: Large-Scale Script Analysis for Cyber Threat Research
  https://arxiv.org/abs/2511.13333v1

• RegionMarker: A Region-Triggered Semantic Watermarking Framework for Embedding-as-a-Service Copyright Protection
  https://arxiv.org/abs/2511.13329v1

• Whistledown: Combining User-Level Privacy with Conversational Coherence in LLMs
  https://arxiv.org/abs/2511.13319v1

• DualTAP: A Dual-Task Adversarial Protector for Mobile MLLM Agents
  https://arxiv.org/abs/2511.13248v1

• A Secure Semantic Communication System Based on Knowledge Graph
  https://arxiv.org/abs/2511.13246v1

• SoK: The Last Line of Defense: On Backdoor Defense Evaluation
  https://arxiv.org/abs/2511.13143v1

• VEIL: Jailbreaking Text-to-Video Models via Visual Exploitation from Implicit Language
  https://arxiv.org/abs/2511.13127v1

• SmartPoC: Generating Executable and Validated PoCs for Smart Contract Bug Reports
  https://arxiv.org/abs/2511.12993v1

• SafeGRPO: Self-Rewarded Multimodal Safety Alignment via Rule-Governed Policy Optimization
  https://arxiv.org/abs/2511.12982v1

• The Grain Family of Stream Ciphers: an Abstraction, Strengthening of Components and New Concrete Instantiations
  https://arxiv.org/abs/2511.12981v1

• Esim: EVM Bytecode Similarity Detection Based on Stable-Semantic Graph
  https://arxiv.org/abs/2511.12971v1

• T2I-Based Physical-World Appearance Attack against Traffic Sign Recognition Systems in Autonomous Driving
  https://arxiv.org/abs/2511.12956v1

• Privacy-Preserving Federated Learning from Partial Decryption Verifiable Threshold Multi-Client Functional Encryption
  https://arxiv.org/abs/2511.12936v1

なお、ポッドキャスト内で紹介する内容は、各論文の概要を日本語で解説したもので、論文概要の著作権は論文著者に帰属します。