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

    Neuronpedia

    INDEX
    Explanations

    programs getting stuck

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     স্ম
    -0.08
     Samuel
    -0.07
    ')[
    -0.07
     ká
    -0.07
    @@
    -0.07
     kapsam
    -0.07
     apreci
    -0.07
    描述
    -0.07
     বিশ্ব
    -0.07
     кое
    -0.07
    POSITIVE LOGITS
     indefinitely
    0.16
     stuck
    0.15
     stalled
    0.14
     долго
    0.13
     无限
    0.12
     અટ
    0.11
     prolonged
    0.11
     slowdown
    0.11
     forever
    0.11
     halted
    0.11
    Act Density 0.008%

    No Known Activations