Glossary term

Activation probes

Techniques used in AI models to monitor internal neuron activations in real-time to detect harmful or misaligned knowledge activation and prevent misuse.

The definition above is the whole entry for this term. Related terms and recent signals are below.