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

    Neuronpedia

    INDEX
    Explanations

    I followed by modal verbs

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     primarily
    0.85
     utilizing
    0.78
     utilizamos
    0.77
    较为
    0.76
     similar
    0.74
     utilizzare
    0.73
    此外
    0.73
     Primarily
    0.72
     utilizar
    0.72
     using
    0.72
    POSITIVE LOGITS
     knew
    1.00
     couldn
    0.99
     feel
    0.90
     shudder
    0.88
    jadi
    0.87
     siente
    0.85
     feels
    0.85
    feel
    0.85
     wouldn
    0.84
     FEEL
    0.81
    Act Density 0.362%

    No Known Activations