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

    Neuronpedia

    INDEX
    Explanations

    negative contractions and the verb "to be."

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

    Followed by a lowercase letter

    np_acts-logits-general · gemini-2.0-flash

    negations like isn't / aren't / won't

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Goodman
    -0.64
    々は
    -0.63
     por
    -0.56
     asemenea
    -0.54
     Persons
    -0.54
     la
    -0.53
     el
    -0.53
    dataSet
    -0.51
     the
    -0.51
     Levinson
    -0.51
    POSITIVE LOGITS
     useAppContext
    1.16
     }}"></
    0.98
     iſt
    0.97
     Walkover
    0.97
     ་་
    0.97
     Anſ
    0.95
    ')";
    0.95
    >>()
    0.93
    />";
    0.93
     Charlemagne
    0.93
    Act Density 0.122%

    No Known Activations