Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    the repetition of the word "keep" and variations indicating maintenance or continuity

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     origin
    -0.51
     AssemblyTitle
    -0.48
     Conven
    -0.47
     telle
    -0.47
     means
    -0.46
    comcast
    -0.46
     perspective
    -0.46
     whiteColor
    -0.46
    ‚¬
    -0.46
     match
    -0.45
    POSITIVE LOGITS
    GraphicsUnit
    1.07
    keep
    0.93
    Keep
    0.90
    kept
    0.90
    rungsseite
    0.85
     lenker
    0.85
     Keep
    0.82
    IVEREF
    0.82
     KEEP
    0.80
    KEEP
    0.80
    Act Density 0.080%

    No Known Activations