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

    Neuronpedia

    INDEX
    Explanations

    proxy socks rules

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    平均
    0.39
    悴
    0.39
    ossing
    0.36
    ングラス
    0.36
    bleeds
    0.36
    blos
    0.36
    oproteins
    0.36
    lair
    0.36
    acid
    0.35
    component
    0.35
    POSITIVE LOGITS
     socks
    0.47
    规则
    0.43
     Regeln
    0.43
     Socks
    0.42
     тол
    0.41
     reglas
    0.40
     Divided
    0.39
     sock
    0.39
     PAC
    0.39
     Sock
    0.39
    Act Density 0.007%

    No Known Activations