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

    Neuronpedia

    INDEX
    Explanations

    references to incomplete or pending tasks and implementations in programming or project contexts

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

    unimplemented features

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     GenerationType
    -0.95
    ItemBackground
    -0.86
    EndContext
    -0.85
     AssemblyCulture
    -0.84
    باردا
    -0.84
    ſſung
    -0.82
    <unused14>
    -0.81
    parsedMessage
    -0.81
     שוליים
    -0.81
    [@BOS@]
    -0.81
    POSITIVE LOGITS
    …
    0.38
     t
    0.37
    ...
    0.37
     (
    0.36
     NotImplemented
    0.35
    t
    0.34
    (
    0.33
    NotImplemented
    0.33
     lot
    0.32
     T
    0.32
    Act Density 0.016%

    No Known Activations