Name: Your LLM Deception Monitor Is Broken. The Fix Is in the Training Data - Sachin Kumar, LexisNexis
Uploaded: 2026-07-08
Description: You fine-tune LLMs and ship them. Your evals are green, your behavioral monitors are green — and a sleeper-agent backdoor can still flip the model to harmful output on a trigger you never tested. Behavioral testing can't reach it, and the interpretability tool people reach for — joint cross-model features (crosscoders)

Why it matters

You fine-tune LLMs and ship them. Your evals are green, your behavioral monitors are green — and a sleeper-agent backdoor can still flip the model to harmful output on a trigger you never tested. Behavioral testing can't reach it, and the interpretability tool people reach for — joint cross-model features (crosscoders)

My takeaway: Your LLM Deception Monitor Is Broken. The Fix Is in the Training Data - Sachin Kumar, LexisNexis is an agent-security signal. The practical read is that autonomy, memory, tool permissions, and third-party integrations are the control surface that needs threat modeling and monitoring.