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

    Neuronpedia

    INDEX
    Explanations

    instances of the word "collapse"

    oai_token-act-pair · gpt-3.5-turbo

    references to structural failures or collapses

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    lean
    -0.80
    etch
    -0.79
    qua
    -0.77
    ju
    -0.77
    vag
    -0.75
    cil
    -0.74
    yah
    -0.72
    zee
    -0.71
    nda
    -0.69
    onga
    -0.68
    POSITIVE LOGITS
    ulence
    0.84
     containment
    0.78
    opian
    0.78
    hower
    0.77
    ateral
    0.76
     inertia
    0.75
    icter
    0.75
     dism
    0.70
    ulent
    0.70
     asleep
    0.70
    Act Density 0.058%

    No Known Activations