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

    Neuronpedia

    INDEX
    Explanations

    humans do

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Doing
    0.96
     Doing
    0.96
    doing
    0.87
     doing
    0.78
    做了
    0.73
     করেননি
    0.64
     робити
    0.64
    去做
    0.64
    要做
    0.63
     melakukan
    0.61
    POSITIVE LOGITS
     does
    1.08
     did
    1.07
     do
    0.96
    does
    0.81
    did
    0.79
    Did
    0.65
    Does
    0.62
    do
    0.60
     Does
    0.59
     Did
    0.58
    Act Density 0.028%

    No Known Activations