Author name: Realmbird

In my previous post I found that activation steering worked with KV_cache and not with hidden state steering.So I decided to look at the PCA with methods such as logit lens and activation steeringQuick Summary:PCA 1 from hidden state activations strong…

In my previous post I found evidence consistent with the scratchpad paper’s compute/store alternation hypothesis — even steps showing higher intermediate answer detection and odd steps showing higher entropy along with results matching “Can we interpre…