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

    Neuronpedia

    INDEX
    Explanations

    titles and subsequent words

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
      
    0.61
    <start_of_image>
    0.54
     using
    0.45
    ..
    0.44
     (
    0.44
     these
    0.43
     member
    0.42
       
    0.42
     depending
    0.41
     here
    0.41
    POSITIVE LOGITS
    <unused1954>
    0.76
    <unused162>
    0.76
    <unused1834>
    0.76
    <unused1153>
    0.76
    <unused1992>
    0.74
    <unused1678>
    0.73
    <unused305>
    0.73
    <unused1845>
    0.73
    <unused565>
    0.72
    <unused1101>
    0.71
    Act Density 0.029%

    No Known Activations