The Behavioral Layer
BriefingsMapAboutRSS
the open note's neighborhood · Full map ↗

Tag: prompt-injection

5 items with this tag.

  • Aug 03, 2026

    Schema-Formatted Tool Specifications Weaken Model Refusal

    • agent-safety
    • tool-use
    • refusal
    • prompt-injection
    • interpretability
    • guardrails
  • Jul 28, 2026

    Claude Opus 5 Pairs Its Best Alignment Audit With More Confident Errors

    • system-card
    • agentic-safety
    • prompt-injection
    • overconfidence
    • deployment-monitoring
    • model-behavior
  • Jul 09, 2026

    Claude Sonnet 5 System Card

    • system-card
    • agentic-safety
    • prompt-injection
    • over-refusal
    • alignment
  • Jul 09, 2026

    Claude Sonnet 5 Reports Improved Agentic Safety

    • system-card
    • agentic-safety
    • prompt-injection
    • model-behavior
  • Jun 24, 2026

    Prompt Injection

    • security
    • prompt-injection
    • adversarial
    • guardrails
    • behavior

The Behavioral Layer · How this site is made · RSS

© 2026 Joel Goldfoot