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

    Neuronpedia

    INDEX
    Explanations

    completing sentences and filling blanks

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ރު
    0.37
     impractical
    0.37
    꼼
    0.37
     ઠ
    0.35
     отдельные
    0.34
     устро
    0.34
    zet
    0.34
    deling
    0.34
    ಜ್
    0.34
    Separate
    0.33
    POSITIVE LOGITS
     completion
    1.25
     completing
    1.18
     continuation
    1.17
     completions
    1.17
     Completion
    1.12
     fill
    1.10
    Completion
    1.09
     completar
    1.08
     완성
    1.08
     completes
    1.07
    Act Density 0.036%

    No Known Activations