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

    Neuronpedia

    INDEX
    Explanations

    instances of "and" used to connect phrases or numbers in lists

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

    "[Number] and a half"

    np_acts-logits-general · gemini-2.0-flash

    quantities involving half

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     communs
    -0.63
    はじめに
    -0.54
     comunale
    -0.48
    MLLoader
    -0.47
     automatiques
    -0.47
    diging
    -0.47
     anyway
    -0.46
     jedenfalls
    -0.46
    limus
    -0.46
    silly
    -0.45
    POSITIVE LOGITS
     half
    0.83
    addCriterion
    0.72
     sundry
    0.67
    Geplaatst
    0.63
     autorytatywna
    0.61
    Half
    0.61
     Half
    0.61
     CreateTagHelper
    0.60
    sund
    0.60
    half
    0.60
    Act Density 0.084%

    No Known Activations