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

    Neuronpedia

    INDEX
    Explanations

    `net` dropped, `process` upgrades, `sign`ificantly

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ச்சர்
    0.54
    क्त
    0.53
    ن
    0.53
    Oi
    0.51
    Approx
    0.51
    ⌵
    0.51
    ંગી
    0.50
    艨
    0.49
    軼
    0.49
    ш
    0.48
    POSITIVE LOGITS
     in
    0.63
     hand
    0.55
     served
    0.52
     stayed
    0.52
     searched
    0.49
     korea
    0.49
     July
    0.48
     paper
    0.48
     Korean
    0.48
     baked
    0.48
    Act Density 0.004%

    No Known Activations