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

    Neuronpedia

    INDEX
    Explanations

    assignments or initializations in code

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

    variable assignment

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     betweenstory
    -1.33
    RegressionTest
    -1.23
    [@BOS@]
    -1.21
    <unused14>
    -1.21
    <unused52>
    -1.21
    <unused74>
    -1.21
    <unused41>
    -1.21
    <unused28>
    -1.21
    <unused16>
    -1.21
    <unused23>
    -1.21
    POSITIVE LOGITS
     =
    0.23
     either
    0.22
     and
    0.20
    ↵↵
    0.20
     sõ
    0.17
    (
    0.17
    ↵
    0.17
     ujarnya
    0.17
     sæ
    0.16
    .
    0.16
    Act Density 0.003%

    No Known Activations