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

    Neuronpedia

    INDEX
    Explanations

    moves beyond abstraction

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Когда
    1.08
     اگه
    1.06
     或者
    1.06
    They
    1.02
    Cuando
    1.01
     ആണ്
    1.00
     근데
    0.99
     ആയിരുന്നു
    0.98
     最近
    0.98
     они
    0.97
    POSITIVE LOGITS
     anxieties
    1.00
     notions
    0.99
     skepticism
    0.91
     cynicism
    0.90
     idealism
    0.87
     culp
    0.87
     humanity
    0.86
     subtlety
    0.85
     perceptions
    0.84
     resentment
    0.84
    Act Density 0.205%

    No Known Activations