Tag: explainability

Showing 1-2 of 2

May 11, 20266 min readModel Release BriefIntermediate

Why LLMs hallucinate — product fixes: triage, grounding and monitoring

Quick summary of the explainer video on why LLMs produce confident-but-false answers, with a practical checklist: verify outputs, add triage, grounding and monitoring before shipping.

hallucination explainability video machine-learning product-management

+1 more

risk-management

Model BreakdownsFrance

Open

Feb 06, 20267 min readFounder NotesIntermediate5 min build

Adversarial Explanation Attacks: How LLM Framing Preserves User Trust in Incorrect Outputs

Describes 'adversarial explanation attacks'—how LLM explanation framing keeps users trusting incorrect outputs. Reports a 205‑participant study and gives pragmatic builder controls.

ai-safety explainability adversarial trust llms

+2 more

product founder