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

    Neuronpedia

    INDEX
    Explanations

    clarifying assumptions

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
    yeah
    -0.10
     eveneens
    -0.08
    again
    -0.08
    そして
    -0.08
     passat
    -0.08
    続
    -0.08
    ју
    -0.08
    /*↵
    -0.08
    -0.08
    aneng
    -0.08
    POSITIVE LOGITS
     restrict
    0.09
     constrain
    0.09
     constraints
    0.09
     impose
    0.08
     restricting
    0.08
     physi
    0.08
     constraint
    0.08
     restrictive
    0.08
     unrealistic
    0.08
     preconce
    0.07
    Act Density 0.050%

    No Known Activations