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

    Neuronpedia

    INDEX
    Explanations

    common articles/pronouns

    np_max-act · gemini-2.0-flash

    This neuron activates on the word “In” when it begins a new sentence or paragraph, marking sentence‐initial discourse transitions.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Sharing
    -0.07
     Syn
    -0.07
     relax
    -0.06
    eding
    -0.06
     Play
    -0.06
    	sd
    -0.06
    asting
    -0.06
    	n
    -0.06
    Playable
    -0.06
     Spot
    -0.06
    POSITIVE LOGITS
    hait
    0.07
    む
    0.07
    (lambda
    0.07
     not
    0.07
    乌
    0.06
     ;;↵
    0.06
     Fortunately
    0.06
     puede
    0.06
    Translatef
    0.06
     pla
    0.06
    Act Density 0.219%

    No Known Activations