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

    Neuronpedia

    INDEX
    Explanations

    same instructions, prompt, RTX, core, description

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    is
    1.30
    न
    1.23
    ן
    1.17
    skog
    1.13
    EN
    1.12
     occidental
    1.10
     Excellency
    1.08
    нь
    1.07
    AL
    1.06
    が増
    1.02
    POSITIVE LOGITS
    एक
    1.21
    ことが多い
    1.12
    𝐒
    1.10
    轍
    1.09
     ተመሳሳይ
    1.08
    𝐍
    1.07
    ͝
    1.05
    ב
    1.03
    ఒ
    1.03
     वही
    1.01
    Act Density 0.021%

    No Known Activations