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

    Neuronpedia

    INDEX
    Explanations

    introducing a summary table

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     persists
    0.82
     belongs
    0.67
     diminishes
    0.67
     requires
    0.65
     worsens
    0.65
     occurs
    0.65
     interferes
    0.64
     originates
    0.63
     fluctuates
    0.62
     пи
    0.62
    POSITIVE LOGITS
     Recap
    0.86
     Summar
    0.86
     Wrap
    0.85
     Wrapping
    0.84
     Consumer
    0.79
     Resolution
    0.78
    ủng
    0.77
    Consumer
    0.76
     調
    0.74
    网友
    0.73
    Act Density 0.073%

    No Known Activations