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

    Neuronpedia

    INDEX
    Explanations

    ensuring correctness

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    en
    1.04
    せて
    1.02
    likely
    0.99
    BOOL
    0.97
     likely
    0.94
     needing
    0.93
     outright
    0.93
    োয়া
    0.89
     devrait
    0.89
    ᐟ
    0.89
    POSITIVE LOGITS
     adequate
    1.34
    足够的
    1.34
     ausreiche
    1.21
    adequate
    1.14
     genü
    1.14
     соблю
    1.11
     충분
    1.10
     proper
    1.10
     Adequate
    1.09
     complies
    1.07
    Act Density 0.066%

    No Known Activations