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

    Neuronpedia

    INDEX
    Explanations

    emphasis and punctuation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     versions
    0.41
     hinges
    0.39
     graces
    0.38
     enhancements
    0.38
     features
    0.37
     incorporates
    0.36
     worksheets
    0.36
     consists
    0.36
     distinguishes
    0.36
     disbursement
    0.35
    POSITIVE LOGITS
     весьма
    0.89
     досить
    0.80
     wow
    0.79
     Wow
    0.77
     Sounds
    0.76
     Hmm
    0.75
     довольно
    0.74
     খুবই
    0.74
     상당히
    0.73
     hmm
    0.73
    Act Density 0.019%

    No Known Activations