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

    Neuronpedia

    INDEX
    Explanations

    check condition

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Often
    1.05
     нередко
    1.05
     spesso
    1.04
     Often
    1.02
     अक्सर
    1.02
     आमतौर
    1.02
     সাধারণত
    1.01
    often
    1.00
     often
    0.99
     geralmente
    0.95
    POSITIVE LOGITS
     exists
    1.30
     meets
    1.25
     exceeds
    1.02
     survives
    0.95
     meet
    0.94
     exist
    0.94
     fails
    0.92
     hasn
    0.91
     existed
    0.91
     matches
    0.90
    Act Density 0.491%

    No Known Activations