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

    Neuronpedia

    INDEX
    Explanations

    phrases related to reasoning or explanation

    oai_token-act-pair · gpt-3.5-turbo

    instances of the words "this" and "that" along with their variations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    Cho
    -0.72
    venants
    -0.70
    lets
    -0.68
    \\\\\\\\
    -0.66
    leon
    -0.66
    mare
    -0.64
    marks
    -0.64
    fl
    -0.63
    ãĤ§
    -0.63
    ãĥ³ãĤ¸
    -0.62
    POSITIVE LOGITS
     sake
    1.67
     purposes
    1.38
     purpose
    1.07
     reasons
    1.02
    ummies
    1.01
     reason
    0.89
     occasion
    0.86
     foreseeable
    0.85
     particular
    0.80
     upcoming
    0.78
    Act Density 0.067%

    No Known Activations