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

    Neuronpedia

    INDEX
    Explanations

    perform actions

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ަށް
    0.43
    ائد
    0.40
     prüfe
    0.40
    Wrong
    0.39
     Wrong
    0.39
    zantine
    0.39
     paraissent
    0.39
     Verificar
    0.38
    刪
    0.38
     അധികാര
    0.38
    POSITIVE LOGITS
     effectively
    0.85
     nicely
    0.77
     handle
    0.76
     achieve
    0.75
     efficiently
    0.73
     provide
    0.73
     handles
    0.72
     capture
    0.67
     establish
    0.65
     accommodate
    0.65
    Act Density 0.658%

    No Known Activations