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

    Neuronpedia

    INDEX
    Explanations

    Key improvements and explanations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     disadvantage
    0.46
     corrected
    0.44
     correcting
    0.42
     verification
    0.42
     fastest
    0.41
     dangerous
    0.40
     dominant
    0.39
     proving
    0.39
     budgeting
    0.39
     unified
    0.39
    POSITIVE LOGITS
    Key
    0.92
     Key
    0.78
     explanations
    0.78
    Explanation
    0.77
     key
    0.75
    key
    0.73
    KEY
    0.70
     Explanation
    0.68
    Explain
    0.66
    Highlights
    0.65
    Act Density 0.021%

    No Known Activations