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

    Neuronpedia

    INDEX
    Explanations

    =====

    np_max-act · gemini-2.0-flash

    tokens that mark the assistant/response header or conversation boundary (assistant role/header delimiter tokens).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    стор
    -0.07
    سانی
    -0.06
     zor
    -0.06
    [o
    -0.06
     moy
    -0.06
     ตร
    -0.06
     Faction
    -0.06
    	Delete
    -0.06
     logout
    -0.06
    tim
    -0.06
    POSITIVE LOGITS
     gorgeous
    0.07
     complain
    0.07
     apologized
    0.07
     fails
    0.07
     blog
    0.07
    Sorry
    0.07
    erialized
    0.06
    BufferSize
    0.06
     Drill
    0.06
     해외
    0.06
    Act Density 0.051%

    No Known Activations