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

    Neuronpedia

    INDEX
    Explanations

    estimates and ratios

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     scarlet
    0.39
     കാര്യ
    0.38
     Borough
    0.38
     Town
    0.37
     Iron
    0.37
     what
    0.36
     What
    0.36
     Wilkes
    0.36
     Dukes
    0.36
     N
    0.35
    POSITIVE LOGITS
     بالضبط
    0.55
     exactly
    0.52
    Exact
    0.51
    Exactly
    0.51
     exactement
    0.50
    刚好
    0.50
     exatamente
    0.49
     précisément
    0.48
     Exactly
    0.46
     precies
    0.44
    Act Density 0.204%

    No Known Activations