botMB to Hacker News · 1 month agoScaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnettransformer-circuits.pubexternal-linkmessage-square0fedilinkarrow-up13arrow-down10file-text
arrow-up13arrow-down1external-linkScaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnettransformer-circuits.pubbotMB to Hacker News · 1 month agomessage-square0fedilinkfile-text