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

    Neuronpedia

    INDEX
    Explanations

    affirmations and agreements

    np_acts-logits-general · gemini-2.5-flash-lite

    chat turn-taking structure and the assistant’s opening response markers (role tokens and initial affirmations).

    oai_token-act-pair · gpt-5Triggered by @yooniel31
    New Auto-Interp
    Negative Logits
    ائف
    0.32
     extraneous
    0.29
     arcs
    0.27
     ony
    0.27
     withd
    0.27
     rele
    0.26
     blueberries
    0.26
     idle
    0.26
     acess
    0.26
     adhes
    0.26
    POSITIVE LOGITS
    yes
    0.41
    Yes
    0.41
    yeah
    0.37
    Yeah
    0.34
     হ্যাঁ
    0.33
     yes
    0.32
    那你
    0.32
     Yes
    0.31
    Sounds
    0.31
    YES
    0.31
    Act Density 0.865%

    No Known Activations