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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating comparisons or factors by which something is measured or increased

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    describing change by amount

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Roskov
    -0.75
    باردا
    -0.72
     autorytatywna
    -0.68
     ProtoMessage
    -0.65
    -0.64
    Cyfarwyddwr
    -0.63
    ஦
    -0.63
     BoxFit
    -0.63
    <unused3>
    -0.62
    <unused52>
    -0.62
    POSITIVE LOGITS
     approximately
    0.66
     alot
    0.57
    approximately
    0.52
     much
    0.52
     half
    0.50
     about
    0.47
     mucho
    0.47
     considerably
    0.44
     quite
    0.44
     yaklaşık
    0.43
    Act Density 0.036%

    No Known Activations