x

LESSWRONG

LW

Mpofu Professor — LessWrong

Mpofu Professor

Mpofu Professor

Message

1

2mo

Mpofu Professor

2mo

Does Social Pressure Leaves A Fingerprint Inside GPT-2? A Controlled Probing Study.

Sharing for critique and feedback. This a single author ,early stage , independent research. TL,DR I carried out a linear probes training on residual stream activations of GPT-2 small and Pythia-160M with in mind " Does prompts framed with social pressure leave a detectable internal signature that is different from...

Sycophancy Under Pressure: A Small ,Controlled Probing Study on Open Models

A bit of personal context:I am doing this on an 8gb RAM windows 10, working mostly on powershell with a goal of building real, demostrable interpretability work without formal ML background , If anything below looks beginner level, it probably is. TL;DR I trained linear probes on the residual stream...