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

    Neuronpedia

    INDEX
    Explanations

    AI refusal or disclaimer

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     dig
    0.42
     cs
    0.38
    ছা
    0.36
     ssl
    0.35
     backwards
    0.35
     stack
    0.34
     damned
    0.34
    ちゃんと
    0.34
     pegs
    0.34
     esper
    0.34
    POSITIVE LOGITS
    Sorry
    0.60
    sorry
    0.58
    Disclaimer
    0.57
    This
    0.56
    Dear
    0.56
     Sorry
    0.56
     Disclaimer
    0.52
    मैं
    0.51
     cautioned
    0.51
     sorry
    0.49
    Act Density 0.013%

    No Known Activations