bot@lemmy.smeargle.fansMB to Hacker News@lemmy.smeargle.fans · 5 months agoScaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnettransformer-circuits.pubexternal-linkmessage-square0fedilinkarrow-up11arrow-down10file-text
arrow-up11arrow-down1external-linkScaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnettransformer-circuits.pubbot@lemmy.smeargle.fansMB to Hacker News@lemmy.smeargle.fans · 5 months agomessage-square0fedilinkfile-text