Detecting High-Stakes Interactions with Activation Probes
Alex McKenzie, Urja Pawar, Phil Blandfort, William Bankes, David Krueger, Ekdeep Singh Lubana, … (+1 more) — 2025-06-12 — arXiv
Summary
Develops and evaluates activation probe architectures for detecting high-stakes LLM interactions that might lead to significant harm, demonstrating robust generalization from synthetic training data to real-world data with six orders-of-magnitude computational savings over LLM-based monitors.
Key Result
Probes trained on synthetic data achieve performance comparable to prompted or finetuned medium-sized LLM monitors while offering computational savings of six orders-of-magnitude.
Source
- Link: https://arxiv.org/abs/2506.10805
- Listed in the Shallow Review of Technical AI Safety 2025 under 1 agenda(s):
- lie-and-deception-detectors — White-box safety (i.e. Interpretability)