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

    Neuronpedia

    INDEX
    Explanations

    evaluation criteria and metrics

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    candle
    0.57
     Siege
    0.57
     DPI
    0.57
     transition
    0.56
    qy
    0.56
    料
    0.54
     mindset
    0.54
    呵
    0.54
    vmin
    0.54
     pouring
    0.54
    POSITIVE LOGITS
     completeness
    1.17
     Relevance
    1.10
     originality
    1.06
     timeliness
    0.94
     uniqueness
    0.93
    Comple
    0.92
     relevance
    0.92
     Comple
    0.92
    Frequency
    0.88
    levance
    0.86
    Act Density 0.375%

    No Known Activations