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

    Neuronpedia

    INDEX
    Explanations

    words related to criticisms and evaluations

    oai_token-act-pair · gpt-3.5-turbo

    elements relating to criticism or negative aspects of experiences

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    atari
    -0.53
    agen
    -0.53
    amera
    -0.50
    pload
    -0.50
    ptives
    -0.49
     Nationwide
    -0.49
     sidx
    -0.48
    ortmund
    -0.48
    acas
    -0.47
    plings
    -0.47
    POSITIVE LOGITS
     nonetheless
    0.66
     underpin
    0.56
     characteristic
    0.56
     nevertheless
    0.56
     behav
    0.55
    Ability
    0.55
    etheless
    0.54
    including
    0.53
     outweigh
    0.53
     inherent
    0.53
    Act Density 1.677%

    No Known Activations