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

    Neuronpedia

    INDEX
    Explanations

    phrases denoting directions or specific locations

    oai_token-act-pair · gpt-3.5-turbo

    hyphenated phrases or clauses, particularly those emphasizing a connection or continuation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ysis
    -0.74
    hou
    -0.68
    ipop
    -0.67
    rons
    -0.67
     cons
    -0.67
    pper
    -0.66
     basil
    -0.65
     lag
    -0.64
     butterflies
    -0.64
    ĵĺ
    -0.64
    POSITIVE LOGITS
    _-
    1.01
    avanaugh
    0.76
    why
    0.76
    [[
    0.75
    feat
    0.75
    yes
    0.74
    ->
    0.74
    âĸº
    0.72
    DERR
    0.71
    tags
    0.71
    Act Density 0.041%

    No Known Activations