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

    Neuronpedia

    INDEX
    Explanations

    specific events and details

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Optimize
    0.49
    高效
    0.47
     optimized
    0.46
     Debug
    0.45
     Used
    0.44
    Designed
    0.44
    Naturally
    0.44
    効率
    0.43
    強制
    0.43
     Designed
    0.43
    POSITIVE LOGITS
     specifics
    0.53
     específica
    0.51
     konkrét
    0.51
     particulars
    0.50
     específicas
    0.48
     envolvendo
    0.48
     específico
    0.47
     menurut
    0.46
     modalités
    0.46
     yanında
    0.45
    Act Density 0.274%

    No Known Activations