本日紹介した論文の一覧 SteerDiff: Steering towards Safe Text-to-Image Diffusion Models http://arxiv.org/abs/2410.02710v1 Discovering Clues of Spoofed LM Watermarks http://arxiv.org/abs/2410.02693v1 Agent Security Bench (ASB): Formalizing and Benchmarking Attacks and Defenses in LLM-based Agents http://arxiv.org/abs/2410.02644v1 Inapproximability of Sparsest Vector in a Real Subspace http://arxiv.org/abs/2410.02636v1 Assessing the Viability of Synthetic Physical Copy Detection Patterns on Different Imaging Systems http://arxiv.org/abs/2410.02575v1 Exploiting HDMI and USB Ports for GPU Side-Channel Insights http://arxiv.org/abs/2410.02539v1 An Edge-Computing based Industrial Gateway for Industry 4.0 using ARM TrustZone Technology http://arxiv.org/abs/2410.02529v1 Encryption-Friendly LLM Architecture http://arxiv.org/abs/2410.02486v1 Towards a Theoretical Understanding of Memorization in Diffusion Models http://arxiv.org/abs/2410.02467v1 Towards a Self-rescuing System for UAVs Under GNSS Attack http://arxiv.org/abs/2410.02442v1 Optimizing Adaptive Attacks against Content Watermarks for Language Models http://arxiv.org/abs/2410.02440v1 Jailbreak Antidote: Runtime Safety-Utility Balance via Sparse Representation Adjustment in Large Language Models http://arxiv.org/abs/2410.02298v1 Alignment of Cybersecurity Incident Prioritisation with Incident Response Management Maturity Capabilities http://arxiv.org/abs/2410.02259v1 MTDNS: Moving Target Defense for Resilient DNS Infrastructure http://arxiv.org/abs/2410.02254v1 Mitigating Downstream Model Risks via Model Provenance http://arxiv.org/abs/2410.02230v1 The Role of piracy in quantum proofs http://arxiv.org/abs/2410.02228v1 Buckle Up: Robustifying LLMs at Every Customization Stage via Data Curation http://arxiv.org/abs/2410.02220v1 BACKTIME: Backdoor Attacks on Multivariate Time Series Forecasting http://arxiv.org/abs/2410.02195v1 BadCM: Invisible Backdoor Attack Against Cross-Modal Learning http://arxiv.org/abs/2410.02182v1 Controlled Generation of Natural Adversarial Documents for Stealthy Retrieval Poisoning http://arxiv.org/abs/2410.02163v1 RiskSEA : A Scalable Graph Embedding for Detecting On-chain Fraudulent Activities on the Ethereum Blockchain http://arxiv.org/abs/2410.02160v1 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。]]

Ask this episode

“What did the council decide on item 42?” — answered with the speaker and the moment.

Transcript

Related