How One Attention Head Acts as a Switch to Amplify Tiny Probe Differences
Sauers_ · x · 2026-10-18
Sauers explains the mechanism: the detection head acts like a switch — without the probe it attends to </think> and pushes toward "No"; with the probe it attends to the user token and pushes toward "Yes." Layers 34-35 use cancellation (like zeroing a scale) to amplify the probe's tiny difference in the user token's key, which normalization turns into a large change of attention target — and training has wired that target to make the model say yes.
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