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

    Neuronpedia

    INDEX
    Explanations

    instances of violent actions and their descriptions

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

    punching and attacking

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     FetchType
    -0.32
    ikbaar
    -0.29
    dientemente
    -0.27
    需要
    -0.27
    tabilité
    -0.26
    ibilidade
    -0.26
    </tfoot>
    -0.26
    hörigen
    -0.24
    ność
    -0.24
    productId
    -0.24
    POSITIVE LOGITS
     punches
    0.71
    punch
    0.68
     attack
    0.66
     punching
    0.64
     hitting
    0.63
    attack
    0.62
    Punch
    0.61
     attacking
    0.60
     enfans
    0.59
     shots
    0.59
    Act Density 0.201%

    No Known Activations