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

    Neuronpedia

    INDEX
    Explanations

    occurrences of whitespace or specific formatting characters

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

    press-and-hold, pull-downs

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    <unused62>
    -0.32
    <unused61>
    -0.31
     począ
    -0.27
     latar
    -0.26
    ↵↵
    -0.25
    <unused60>
    -0.25
     irse
    -0.24
    stø
    -0.23
     blandt
    -0.23
    <unused63>
    -0.23
    POSITIVE LOGITS
    OGND
    1.40
     AssemblyCulture
    1.05
    <unused52>
    1.05
    <unused74>
    1.05
    <unused28>
    1.05
    <unused8>
    1.05
    <unused41>
    1.05
    <unused68>
    1.05
    <unused43>
    1.05
    <pad>
    1.04
    Act Density 0.008%

    No Known Activations