Skip to content
CBContext/Bleed Wiki
← Evidence library
Preprint

Research · 2026

From Untrusted Input to Trusted Memory

Dash et al. · arXiv

What it says

Studies how a single adversarial memory write can retain influence across later agent interactions.

Why it matters here

Distinguishes malicious persistent-memory corruption from accidental context carryover.

Editorial caution

This record has not completed peer review. Treat terminology, measurements and generalisations as provisional.

memorypoisoningagents
Open original source