PREVIEW. This is how it will look in an inbox; it was not sent to you. Each real email carries its own unsubscribe link and Keryn's postal address.

Keryn Headlines

Daily, Monday 19 May 2025 · Retrospective · Covers 18 May 2025 00:00 to 19 May 2025 00:00 UTC

Read on the web

Top developments

A Survey of Attacks on Large Language Models

arXiv · Research · Models

SPIRIT: Patching Speech Language Models against Jailbreak Attacks

arXiv · Research · Models, Remediation

Improving Google A2A Protocol: Protecting Sensitive Data and Mitigating Unintended Harms in Multi-Agent Systems

arXiv · Research · AI Security, Models

Agentic SDLC

EVALOOOP: A Self-Consistency-Centered Framework for Assessing Large Language Model Robustness in Programming

arXiv · Research · Agentic SDLC, Models

Models

Improving LLM Outputs Against Jailbreak Attacks with Expert Model Integration

arXiv · Research · AI Security, Models

AI Security

CAPTURE: Context-Aware Prompt Injection Testing and Robustness Enhancement

arXiv · Research · AI Security

Titles, links and short excerpts only; every item links to its source. Selection is automatic; the editor's note is AI-drafted and checked as described. Preview: the unsubscribe link and Keryn's postal address are added to each email when it is sent.