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

    Neuronpedia

    INDEX
    Explanations

    words related to comparisons and contrasts

    oai_token-act-pair · gpt-3.5-turbo

    repetitive phrases related to contrasting viewpoints or sides in discussions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    bis
    -0.72
    weights
    -0.70
    packages
    -0.70
    inals
    -0.68
    cells
    -0.66
    riages
    -0.65
    ée
    -0.65
    iaries
    -0.65
    Magikarp
    -0.63
    redients
    -0.62
    POSITIVE LOGITS
     occasion
    1.02
     occasions
    0.95
     periphery
    0.75
     heels
    0.74
     contrary
    0.74
     Scene
    0.71
     eve
    0.69
     Janeiro
    0.68
     assumption
    0.68
    Ĥİ
    0.67
    Act Density 0.116%

    No Known Activations