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

    Neuronpedia

    INDEX
    Explanations

    quotation marks

    np_max-act · gemini-2.0-flash

    This neuron responds to occurrences of the first‐person pronoun “I” (including forms like “I’m”) in dialogue.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    dialogue lines in quoted speech, especially openings and first‑person pronoun/contraction patterns within the utterance.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     Vapor
    -0.07
    (progress
    -0.07
    finity
    -0.07
     исч
    -0.06
     توجه
    -0.06
     """
    -0.06
    .faces
    -0.06
    ователь
    -0.06
    ropolitan
    -0.06
     получить
    -0.06
    POSITIVE LOGITS
    เม
    0.06
    _COMMON
    0.06
    illegal
    0.06
    СТ
    0.06
    ((_
    0.06
    нівер
    0.06
     unfold
    0.06
    'y
    0.06
     初
    0.06
    _OPTS
    0.06
    Act Density 0.039%

    No Known Activations