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

    Neuronpedia

    INDEX
    Explanations

    variable names starting with my_

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     CDs
    0.97
     DVDs
    0.94
     GPUs
    0.91
     URLs
    0.88
     классов
    0.88
     NPCs
    0.87
     Sections
    0.84
     CPUs
    0.84
     huts
    0.83
    Sections
    0.83
    POSITIVE LOGITS
     equation
    0.99
     theorem
    0.95
     article
    0.93
     quarter
    0.92
     episode
    0.92
     molecule
    0.88
     table
    0.88
    equation
    0.87
     column
    0.86
     book
    0.86
    Act Density 1.205%

    No Known Activations