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

    Neuronpedia

    INDEX
    Explanations

    negation verbs

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    まさに
    0.67
    been
    0.64
    ামুটি
    0.64
     Been
    0.63
     Guilty
    0.62
    又是
    0.62
    egd
    0.62
     вот
    0.62
     been
    0.62
     були
    0.62
    POSITIVE LOGITS
     doesn
    4.81
     didn
    4.34
    doesn
    4.21
     Doesn
    4.11
    Doesn
    3.84
     doesnt
    3.64
     isn
    3.59
     Didn
    3.54
     não
    3.51
    didn
    3.49
    Act Density 1.807%

    No Known Activations