Description

本日紹介した論文の一覧

• Evolution of Cybersecurity Subdisciplines: A Science of Science Study
  https://arxiv.org/abs/2511.19331v1

• Medusa: Cross-Modal Transferable Adversarial Attacks on Multimodal Medical Retrieval-Augmented Generation
  https://arxiv.org/abs/2511.19257v1

• FedPoisonTTP: A Threat Model and Poisoning Attack for Federated Test-Time Personalization
  https://arxiv.org/abs/2511.19248v1

• Adversarial Attack-Defense Co-Evolution for LLM Safety Alignment via Tree-Group Dual-Aware Search and Optimization
  https://arxiv.org/abs/2511.19218v1

• Can LLMs Threaten Human Survival? Benchmarking Potential Existential Threats from LLMs via Prefix Completion
  https://arxiv.org/abs/2511.19171v1

• A General Framework for Per-record Differential Privacy
  https://arxiv.org/abs/2511.19015v1

• Understanding and Mitigating Over-refusal for Large Language Models via Safety Representation
  https://arxiv.org/abs/2511.19009v1

• LLM-CSEC: Empirical Evaluation of Security in C/C++ Code Generated by Large Language Models
  https://arxiv.org/abs/2511.18966v1

• Defending Large Language Models Against Jailbreak Exploits with Responsible AI Considerations
  https://arxiv.org/abs/2511.18933v1

• RoguePrompt: Dual-Layer Ciphering for Self-Reconstruction to Circumvent LLM Moderation
  https://arxiv.org/abs/2511.18790v1

• Re-Key-Free, Risky-Free: Adaptable Model Usage Control
  https://arxiv.org/abs/2511.18772v1

• Evaluation of Real-Time Mitigation Techniques for Cyber Security in IEC 61850 / IEC 62351 Substations
  https://arxiv.org/abs/2511.18748v1

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。