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

    Neuronpedia

    INDEX
    Explanations

    chemical processes

    np_max-act · gemini-2.0-flash

    This neuron specifically activates on the special conversation‐format control tokens (e.g. the <|start_header_id|> / <|end_header_id|> and similar protocol markers).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     northwest
    -0.07
     ब
    -0.07
    .dtd
    -0.06
     pooling
    -0.06
    ANCE
    -0.06
    _PRIORITY
    -0.06
     mitigation
    -0.06
     userManager
    -0.06
    (vis
    -0.06
     WEIGHT
    -0.06
    POSITIVE LOGITS
    <hr
    0.06
    éo
    0.06
    .currentTime
    0.06
     topLevel
    0.06
     Postal
    0.06
    amı
    0.06
     getInput
    0.06
    annya
    0.06
    intersect
    0.06
     Tahoe
    0.06
    Act Density 0.041%

    No Known Activations