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

    Neuronpedia

    INDEX
    Explanations

    movement and state changes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     rhetorical
    0.73
    捗
    0.72
     cynicism
    0.69
    mittedly
    0.68
     banter
    0.68
     rhetoric
    0.67
     camaraderie
    0.67
     dotycz
    0.66
     alienation
    0.64
     konusunda
    0.63
    POSITIVE LOGITS
     downwards
    1.11
     during
    1.07
     horizontally
    1.06
     creating
    1.06
     rapidly
    1.05
     underground
    1.04
     causing
    1.03
     outwards
    1.02
     allowing
    1.02
     thereby
    1.01
    Act Density 0.681%

    No Known Activations