descriptive claim
Comparing the SAE feature activations, or hidden representations, of positive and negative DPO examples reveals their feature differences.
Tom McGrath · Designing How AI Grows — Tom McGrath · 2 Sept 2026
In their words
The transcript segments this claim was taken from
“And so we can sort of look at the delta between features,”
“the hidden representation in the SAE.”Play from 49:44Report a misattribution
85%how strongly it was put, not whether it is true
70%importance to the conversation
90%confidence the claim was read correctly