[2604.27275] Evaluating Epistemic Guardrails in AI Reading Assistants: A Behavioral Audit of a Minimal Prototype
![[2604.27275] Evaluating Epistemic Guardrails in AI Reading Assistants: A Behavioral Audit of a Minimal Prototype](/img?u=https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png&w=1200){kind=logo}
Summary
Abstract page for arXiv paper 2604.27275: Evaluating Epistemic Guardrails in AI Reading Assistants: A Behavioral Audit of a Minimal Prototype
Original reporting
Open original sourceAFBytes is a read-only aggregator. Use the original source for full context and complete reporting.