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

    Neuronpedia

    INDEX
    Explanations

    phrases related to escalating conflicts and violent situations

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.06
    1:0.03
    2:0.13
    3:0.07
    4:0.17
    5:0.05
    6:0.03
    7:0.03
    8:0.16
    9:0.14
    10:0.06
    11:0.02
    Negative Logits
    stone
    -1.33
    prints
    -1.30
    Score
    -1.26
    ked
    -1.25
    maker
    -1.23
    liest
    -1.23
    boys
    -1.20
    builders
    -1.19
    girls
    -1.19
    stones
    -1.18
    POSITIVE LOGITS
     hostilities
    1.72
     ACTIONS
    1.54
     escalation
    1.53
     tensions
    1.51
     confrontation
    1.50
    escal
    1.47
    ricular
    1.40
    ��
    1.40
     escalating
    1.40
     dialogue
    1.38
    Act Density 0.009%

    No Known Activations