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, Saturday 19 October 2024 · Retrospective · Covers 18 October 2024 00:00 to 19 October 2024 00:00 UTC

Read on the web

Top developments

Sabotage evaluations for frontier models

Anthropic · Research · Models

Microsoft Reveals macOS Vulnerability that Bypasses Privacy Controls in Safari Browser

The Hacker News · Analysis · Remediation

Enhancing Prompt Injection Attacks to LLMs via Poisoning Alignment

arXiv · Research · AI Security

SAST

From Solitary Directives to Interactive Encouragement! LLM Secure Code Generation by Natural Language Prompting

arXiv · Research · SAST

Models

Feint and Attack: Attention-Based Strategies for Jailbreaking and Protecting LLMs

arXiv · Research · Models

Backdoored Retrievers for Prompt Injection Attacks on Retrieval Augmented Generation of Large Language Models

arXiv · Research · AI Security, Models

Unlearning Backdoor Attacks for LLMs with Weak-to-Strong Knowledge Distillation

arXiv · Research · Models

Advisories in this window

  • 11 npm advisories · 11 critical
  • 2 NuGet advisories · 1 high · 1 moderate
  • 2 Go advisories · 1 critical · 1 high
  • 1 Maven advisory · 1 moderate

All advisories

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.