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

    Neuronpedia

    INDEX
    Explanations

    phrases related to discovering something interesting

    oai_token-act-pair · gpt-3.5-turbo

    phrases that introduce or emphasize the presence of various subjects

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    obar
    -0.84
    alties
    -0.83
    anism
    -0.82
    Contents
    -0.80
    onge
    -0.74
    eyes
    -0.74
    cest
    -0.74
    always
    -0.74
    hes
    -0.73
    ees
    -0.73
    POSITIVE LOGITS
     strange
    1.14
     bunch
    1.08
     guy
    1.07
     startling
    1.05
     mysterious
    1.05
     bombshell
    1.04
     typo
    1.03
     wonderful
    1.02
     fascinating
    1.02
     revelation
    1.00
    Act Density 0.518%

    No Known Activations