AI security research / concepts
Mechanisms and patterns across the research.
62 Topics · Page 4 of 4
2 matching records on this page
A website can place operational requests in llms.txt, HTTP headers or other retrieved metadata; the failure occurs when an agent treats that lower-trust content as permission for registration, installation or tool execution.
16 explicit connections
Direct edits to model parameters can change deployed behavior without a conventional fine-tune, so confidence requires checkpoint provenance and behavioral regression, not a familiar model name.
5 explicit connections
No matches on this page.