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

    Neuronpedia

    INDEX
    Explanations

    Casual conversation/requests

    np_max-act · gemini-2.0-flash

    tokens that belong to the assistant's generated message or message/metadata markers (i.e., assistant-role and model-generated content).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    *>&
    -0.07
    ることは
    -0.07
    $result
    -0.07
     Ih
    -0.06
     ZIP
    -0.06
    ектив
    -0.06
    .permission
    -0.06
    ໄ
    -0.06
    (^)(
    -0.06
     IsPlainOldData
    -0.06
    POSITIVE LOGITS
    应力
    0.07
    Area
    0.07
     poly
    0.07
     camping
    0.07
     sala
    0.07
     Thư
    0.07
     SALE
    0.07
     deposits
    0.07
    缝隙
    0.07
     winds
    0.07
    Act Density 1.939%

    No Known Activations