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

    Neuronpedia

    INDEX
    Explanations

    marking important moments

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     occasionally
    0.91
     inhibiting
    0.85
     discourage
    0.85
     discouraging
    0.84
     soms
    0.84
     hindering
    0.83
     sometimes
    0.82
     dissu
    0.82
     inhibitory
    0.82
     hinders
    0.82
    POSITIVE LOGITS
     marks
    1.51
     mark
    1.45
     marking
    1.39
     signal
    1.37
     inaugur
    1.33
     signalled
    1.30
     signals
    1.28
    signal
    1.28
     signaled
    1.27
    marks
    1.27
    Act Density 0.431%

    No Known Activations