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

    Neuronpedia

    INDEX
    Explanations

    Escape

    np_max-act · gemini-2.0-flash

    The neuron is highly sensitive to the word “Escape,” especially when it appears as a standalone title or heading.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Id
    -0.08
     shows
    -0.07
     commodity
    -0.06
     Willow
    -0.06
     Bid
    -0.06
     show
    -0.06
     Hill
    -0.06
     bins
    -0.06
     subdir
    -0.06
    087
    -0.06
    POSITIVE LOGITS
     escape
    0.14
     Escape
    0.12
    Escape
    0.10
     escaped
    0.10
     escapes
    0.09
    escape
    0.09
    CAPE
    0.09
    逃
    0.08
     escap
    0.08
    .escape
    0.08
    Act Density 0.009%

    No Known Activations