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

    Neuronpedia

    INDEX
    Explanations

    helps or allows

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     clamping
    0.42
     mampu
    0.41
     élevée
    0.41
    DYNAMIC
    0.40
     способны
    0.40
     tending
    0.39
     upholding
    0.39
     dini
    0.39
     thence
    0.38
     JOURNAL
    0.38
    POSITIVE LOGITS
     hjæl
    0.55
     ae
    0.53
     aer
    0.52
     help
    0.49
     hjäl
    0.48
     encourage
    0.48
     promote
    0.47
     βοη
    0.45
     encourages
    0.44
     mellow
    0.44
    Act Density 0.025%

    No Known Activations