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

    Neuronpedia

    INDEX
    Explanations

    asking why or understanding problems

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     road
    0.73
     d
    0.73
     hostage
    0.66
     |
    0.65
     nan
    0.65
     //
    0.64
     +
    0.63
    !
    0.62
     (
    0.61
     v
    0.61
    POSITIVE LOGITS
    Understanding
    1.15
    Choosing
    1.15
    Finding
    1.13
    Insights
    1.12
    Selecting
    1.06
    Reasons
    1.03
    Characteristics
    1.02
    Advantages
    1.01
    Problems
    1.01
    Why
    1.00
    Act Density 0.000%

    No Known Activations