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

    Neuronpedia

    INDEX
    Explanations

    focus on operation and perception details

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    😟
    0.50
     whitish
    0.50
     brownish
    0.49
     almond
    0.47
     yellowish
    0.46
    0.46
     "..
    0.45
    😃
    0.45
     ช่วย
    0.44
    ueness
    0.44
    POSITIVE LOGITS
     praxis
    0.54
     bastard
    0.52
    抵达
    0.51
     tether
    0.48
     asymptotically
    0.48
     fetish
    0.48
     prophyl
    0.48
     fundamentally
    0.48
     obsol
    0.48
     desperately
    0.47
    Act Density 0.017%

    No Known Activations