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

    Neuronpedia

    INDEX
    Explanations

    sticky patches

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     правило
    0.46
    zone
    0.44
     хе
    0.42
    rule
    0.42
    ifica
    0.40
     akademik
    0.40
     pertanian
    0.40
     rule
    0.39
    为大家
    0.39
     hor
    0.39
    POSITIVE LOGITS
    φαν
    0.43
    ಿಯೂ
    0.43
     titled
    0.43
     ਵਾਲ
    0.42
    թ
    0.42
    夕
    0.42
     তারাও
    0.41
    فير
    0.41
     gifts
    0.40
    스럽
    0.40
    Act Density 0.001%

    No Known Activations