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

    Neuronpedia

    INDEX
    Explanations

    even USB bundle updates inconsistent

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ).
    0.50
    .
    0.46
    phys
    0.42
    )。
    0.42
    );
    0.42
    lessly
    0.41
    cock
    0.41
    isy
    0.41
    ].
    0.41
     détach
    0.40
    POSITIVE LOGITS
    द्री
    0.48
    我自己
    0.45
    σε
    0.44
     సినిమాలు
    0.44
    ৫০
    0.43
     Футбол
    0.42
    损
    0.42
    전
    0.41
    की
    0.41
     ১
    0.40
    Act Density 0.008%

    No Known Activations