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

    Neuronpedia

    INDEX
    Explanations

    input format description

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     leech
    0.39
     Confirm
    0.39
     debugging
    0.38
     damper
    0.38
     Thermodynamic
    0.38
     don
    0.38
     fishing
    0.37
     thermodynamic
    0.37
     moot
    0.37
     litter
    0.36
    POSITIVE LOGITS
     consists
    0.63
     consisted
    0.57
     terdiri
    0.56
    เป็น
    0.53
     consiste
    0.52
     consist
    0.50
     formato
    0.49
     örnek
    0.49
     formatos
    0.49
     عباره
    0.48
    Act Density 0.007%

    No Known Activations