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

    Neuronpedia

    INDEX
    Explanations

    instances of the words "wrong" and "correct," highlighting discussions around accuracy and errors

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    "wrong" or "wrongful"

    np_acts-logits-general · gemini-2.0-flash

    wrongful

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     AssemblyCompany
    -0.78
     GoogleFonts
    -0.61
    LabelTagHelper
    -0.61
     tramonto
    -0.59
    RectangleBorder
    -0.57
     AssemblyTitle
    -0.56
    IntoConstraints
    -0.56
     newBuilder
    -0.55
     AppBundle
    -0.55
    ьере
    -0.53
    POSITIVE LOGITS
     Wrong
    0.74
     wrong
    0.69
     WRONG
    0.68
    ness
    0.66
    Wrong
    0.65
    WRONG
    0.64
    headed
    0.63
    wrong
    0.63
     proper
    0.62
     Proper
    0.62
    Act Density 0.159%

    No Known Activations