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

    Neuronpedia

    INDEX
    Explanations

    modal verbs indicating possibility or future actions

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

    who can / would / should / said

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    migrationBuilder
    -0.41
    المكان
    -0.41
    Autoritní
    -0.40
    sprü
    -0.39
     InputDecoration
    -0.34
     mucho
    -0.33
     way
    -0.32
     Великой
    -0.32
     beaucoup
    -0.32
    SetUp
    -0.32
    POSITIVE LOGITS
     who
    0.73
     Administrativna
    0.70
    Whom
    0.67
     whom
    0.66
     którzy
    0.65
    rungsseite
    0.63
     Whom
    0.63
    Who
    0.60
    whom
    0.59
     quienes
    0.58
    Act Density 0.010%

    No Known Activations