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

    Neuronpedia

    INDEX
    Explanations

    Concluding words

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     moiety
    0.56
     earthenware
    0.54
     −
    0.54
     ją
    0.53
     moieties
    0.52
     eponymous
    0.52
     messing
    0.50
     residue
    0.50
     हिस्से
    0.50
     abelian
    0.50
    POSITIVE LOGITS
    **
    1.34
    Bottom
    1.15
    ---
    1.13
    ***
    1.13
    So
    1.06
    Finally
    1.04
    Essentially
    1.04
    Therefore
    1.02
    Ultimately
    1.02
    Do
    1.02
    Act Density 0.690%

    No Known Activations