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

    Neuronpedia

    INDEX
    Explanations

    negative qualifiers and expressions of disappointment

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

    Follows intensifiers like "so", "too", or "not"

    np_acts-logits-general · gemini-2.0-flash

    flattering, appealing, pleasant

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ValueStyle
    -0.86
     EconPapers
    -0.67
    AntiForgeryToken
    -0.67
    </tfoot>
    -0.64
    almost
    -0.61
    UnsafeEnabled
    -0.58
     Nauk
    -0.56
    woordig
    -0.56
     lenker
    -0.56
    nigh
    -0.55
    POSITIVE LOGITS
     særlig
    0.79
     ideal
    0.68
    rovna
    0.67
     conducive
    0.67
     flattering
    0.66
     appealing
    0.65
     pleasant
    0.62
     favorable
    0.59
     happy
    0.59
     particularly
    0.59
    Act Density 0.268%

    No Known Activations