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

    Neuronpedia

    INDEX
    Explanations

    somewhat cohesive

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     poisoning
    0.43
    rrbracket
    0.43
     paint
    0.42
     died
    0.42
     paintings
    0.42
     medication
    0.41
     pint
    0.41
     médicaments
    0.40
     comenc
    0.40
     pusieron
    0.40
    POSITIVE LOGITS
    足以
    0.43
    反映
    0.42
    应该是
    0.42
    以便
    0.41
    Rationale
    0.41
     завтра
    0.41
    should
    0.39
     должно
    0.38
     الواقع
    0.38
    郤
    0.38
    Act Density 0.002%

    No Known Activations