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

    Neuronpedia

    INDEX
    Explanations

    word

    np_max-act · gemini-2.0-flash

    The neuron is sensitive to the token “words,” i.e. it flags when the prompt is specifying a word-count constraint.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     nargs
    -0.06
    AnimationFrame
    -0.06
     updater
    -0.06
    _layers
    -0.06
     вход
    -0.06
     zpráva
    -0.06
    .fixed
    -0.06
    	ERR
    -0.06
    UPLOAD
    -0.06
     Quang
    -0.06
    POSITIVE LOGITS
    elo
    0.07
     Curse
    0.06
    任
    0.06
    _Session
    0.06
     fond
    0.06
    uli
    0.06
     cruel
    0.06
    _Current
    0.06
     Pods
    0.06
     billig
    0.06
    Act Density 0.001%

    No Known Activations