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

    Neuronpedia

    INDEX
    Explanations

    exit, leave, escape

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Retrieve
    0.38
     Trace
    0.37
     Strict
    0.35
     Rost
    0.33
    สำหรับ
    0.32
     Normalize
    0.32
     Prepare
    0.32
     Validation
    0.32
     retrieving
    0.32
     Wieder
    0.32
    POSITIVE LOGITS
     khỏi
    0.39
    ла
    0.34
    离开
    0.32
    離開
    0.32
     gracefully
    0.31
    ώσεις
    0.31
    icula
    0.30
     politely
    0.30
    сит
    0.30
    ഠ
    0.30
    Act Density 0.078%

    No Known Activations