Probe detection traced to a single attention head reading the user token's hidden state
Sauers_ · x · 2026-10-18
Activation-swap experiments show the model detects probes via tiny differences in the user token's hidden state: swapping activations at that token alone (99% of effect), or one head's output (head 6, layer 35), fully reproduces the behavior change.
More from Research
- TRACE Open-Sourced: Migrating G1 Controller to Full-Size Humanoid H2 in 8 Hours on One GPU — zhengyiluo · 2026-10-18
- Humanoid Robot Funding Tops $3B as Brian Potter Explains the VLA Models Driving Robots — scottleibrand · 2026-10-18
- Terence Tao on 'War Footing' as OpenAI Claims Navier-Stokes Solve Shakes Mathematics — AlexTensor · 2026-10-18
- Univ. of Tokyo paper: why your recorded voice feels eerie — a vocal uncanny valley — ssh4net · 2026-10-18
- MIT's SANDO Drone Planner Avoids Every Moving Obstacle Across 12 Real Flights — Olivier__OG · 2026-10-18
- No-backprop pretraining speedup of 2.3x claimed, inspired by mech interp essay — CatAstro_Piyush · 2026-10-18