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

    Neuronpedia

    INDEX
    Explanations

    well-formed generation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    us
    1.54
    a
    1.53
    AM
    1.49
    اور
    1.47
    াতি
    1.44
    B
    1.40
    K
    1.39
    ان
    1.39
    IM
    1.38
     bahan
    1.37
    POSITIVE LOGITS
     whirl
    1.63
     Arbor
    1.55
     unjust
    1.54
     airy
    1.53
     ionic
    1.53
     misty
    1.52
     thirties
    1.48
     inevitable
    1.48
     ion
    1.47
     hexagon
    1.47
    Act Density 0.006%

    No Known Activations