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

    Neuronpedia

    INDEX
    Explanations

    causal decoder-only models

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     brah
    0.42
     oscillations
    0.39
     drills
    0.39
     secteurs
    0.39
     جذر
    0.39
    kirk
    0.36
    าลัย
    0.36
     கண்டுபிடிக்க
    0.36
    ラの
    0.36
    ্লার
    0.36
    POSITIVE LOGITS
     causal
    0.98
     autore
    0.78
     causality
    0.71
     Caus
    0.68
     LM
    0.66
    Aut
    0.65
     Aut
    0.63
    autore
    0.61
     conditional
    0.60
    LM
    0.59
    Act Density 0.080%

    No Known Activations