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

    Neuronpedia

    INDEX
    Explanations

    phrases or sequences containing special characters, particularly âĢ

    oai_token-act-pair · gpt-3.5-turbo

    instances of the special character sequence ‘âĢ’

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     cyan
    -0.59
     decomp
    -0.57
     scatter
    -0.55
     scattering
    -0.52
     minim
    -0.50
     gib
    -0.49
     shack
    -0.49
     radar
    -0.49
    creen
    -0.47
     gossip
    -0.47
    POSITIVE LOGITS
    ¹
    0.71
    º
    0.69
    į
    0.68
    ISIS
    0.67
    ¬
    0.66
    ı
    0.66
    £
    0.66
    ihad
    0.63
    sure
    0.62
    say
    0.61
    Act Density 0.571%

    No Known Activations