Praneet @praneet.bsky.social · 19/11/2024Oh, wow! That’s alarming. The algorithm isn’t truly open source then. 100
Reposted by PraneetSonia @soniajoseph.bsky.social · 11/10/2024I wrote a post on multimodal interpretability techniques, including sparse feature circuit discovery, exploiting the shared text-image space of CLIP, and training adapters. soniajoseph.ai/multimodal-i... 💜soniajoseph.aiMultimodal interpretability in 2024Multimodal interpretability, from sparse feature circuits with SAEs, to vision transformers leveraging CLIP's shared text-image space. 0344
