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

    Neuronpedia

    INDEX
    Explanations

    then / sequence

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     влия
    0.81
     beeinfl
    0.79
    atuak
    0.75
     undermines
    0.75
     berücksichtigt
    0.74
     влияния
    0.73
     longstanding
    0.73
     influência
    0.73
     totalitarian
    0.73
     illegitimate
    0.72
    POSITIVE LOGITS
    然后
    1.33
     then
    1.23
    然後
    1.20
    then
    1.15
     ثم
    1.15
     然后
    1.15
    その後
    1.13
    Then
    1.10
     затем
    1.09
     سپس
    1.09
    Act Density 1.951%

    No Known Activations