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

    Neuronpedia

    INDEX
    Explanations

    productmaterialfeatures

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Verbose
    0.73
    criptive
    0.70
    Post
    0.63
    あるいは
    0.63
    anything
    0.62
     nějak
    0.62
    either
    0.61
    IOException
    0.61
    その
    0.60
    Vendor
    0.58
    POSITIVE LOGITS
     ensures
    1.64
     allows
    1.58
     ，
    1.52
    ，
    1.50
     ،
    1.42
     protects
    1.41
     zorgt
    1.41
     garantiza
    1.39
     helps
    1.38
     enables
    1.38
    Act Density 0.016%

    No Known Activations