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

    Neuronpedia

    INDEX
    Explanations

    training or trained data

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     calculations
    0.69
     Bohr
    0.67
    logic
    0.64
    Logic
    0.64
    Logical
    0.63
     hermaph
    0.61
     Calculations
    0.61
     Logic
    0.60
    friger
    0.59
     decisions
    0.58
    POSITIVE LOGITS
     training
    2.21
    Training
    2.11
     Training
    2.11
    training
    2.01
    训练
    1.88
     Trainings
    1.86
     treinamento
    1.79
     TRAINING
    1.78
    培训
    1.75
     pelatihan
    1.67
    Act Density 1.653%

    No Known Activations