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

    Neuronpedia

    INDEX
    Explanations

    FI errors

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ın
    0.40
     அல்லது
    0.40
    ession
    0.40
     môžete
    0.39
    ınıza
    0.39
    хів
    0.38
     이상의
    0.38
     lichaam
    0.38
    Computation
    0.38
    ・
    0.38
    POSITIVE LOGITS
    b
    0.55
    p
    0.53
    m
    0.51
    d
    0.47
     OMG
    0.47
     idiots
    0.46
     not
    0.45
     Habs
    0.42
     been
    0.42
     Damn
    0.41
    Act Density 0.332%

    No Known Activations