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

    Neuronpedia

    INDEX
    Explanations

    identifying and understanding actions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     hugely
    1.09
     confirm
    1.02
     massively
    0.99
     assert
    0.89
     superbly
    0.87
     verify
    0.87
     enormously
    0.84
     vastly
    0.84
    Confirm
    0.83
     assertion
    0.83
    POSITIVE LOGITS
    <eos>
    0.87
     auxquelles
    0.81
    -\\
    0.80
     Darüber
    0.79
    aras
    0.75
    夲
    0.74
    幵
    0.71
    enumi
    0.71
     เมื่อ
    0.70
    㧔
    0.70
    Act Density 0.258%

    No Known Activations