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

    Neuronpedia

    INDEX
    Explanations

    behavior, humor, fulfill

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     generalised
    2.40
     leukaemia
    1.91
     customised
    1.81
     emphasise
    1.80
     rumour
    1.77
     fertilisers
    1.75
     honourable
    1.74
     utilise
    1.73
     flavour
    1.72
    ਸ਼
    1.65
    POSITIVE LOGITS
    ي
    1.85
    ти
    1.63
    ()=>{
    1.61
    abend
    1.58
    𝘦
    1.56
    のが
    1.52
    𝒑
    1.50
    𝒊
    1.49
    і
    1.49
    ড়ী
    1.48
    Act Density 0.467%

    No Known Activations