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

    Neuronpedia

    INDEX
    Explanations

    segment durations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     reservation
    0.96
     lot
    0.94
     reag
    0.81
     reservations
    0.80
     festivities
    0.79
     জাতিসঙ্ঘ
    0.78
     defensive
    0.78
    reserv
    0.77
     concili
    0.76
     lightening
    0.76
    POSITIVE LOGITS
    3
    1.12
    2
    1.11
    4
    1.09
    5
    1.09
    7
    1.05
    6
    1.05
    9
    1.04
    1
    1.00
    8
    0.95
    Solving
    0.93
    Act Density 0.002%

    No Known Activations