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

    Neuronpedia

    INDEX
    Explanations

    Chinese verbs

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     vários
    0.65
     sendo
    0.64
     cibo
    0.63
     sarebbero
    0.63
    过来的
    0.63
     angiel
    0.62
     estavam
    0.59
     possibilidades
    0.59
     tendo
    0.58
     gehad
    0.58
    POSITIVE LOGITS
    向
    0.92
    将
    0.86
    把
    0.76
    將
    0.75
    在
    0.70
    将其
    0.70
    从
    0.69
    利用
    0.69
    进行
    0.68
    用
    0.68
    Act Density 0.064%

    No Known Activations