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

    Neuronpedia

    INDEX
    Explanations

    lessons or experiences that the text discusses learning from

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    whe
    -0.38
    few
    -0.34
    oooooooo
    -0.34
     stret
    -0.33
    eri
    -0.32
    tone
    -0.31
    head
    -0.31
    oooo
    -0.31
    ooo
    -0.30
    home
    -0.30
    POSITIVE LOGITS
     Learned
    0.46
     firsthand
    0.41
    rawdownloadcloneembedreportprint
    0.38
    learn
    0.38
     glean
    0.37
    ĸļ
    0.37
     learnt
    0.36
    fulness
    0.35
    Lear
    0.35
     learned
    0.35
    Act Density 8.347%

    No Known Activations