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

    Neuronpedia

    INDEX
    Explanations

    bash commands

    np_max-act · gemini-2.0-flash

    This neuron responds to code‐formatting tokens (e.g. backticks and fence markers) rather than ordinary words.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     wont
    -0.08
     complement
    -0.07
     }}↵↵
    -0.07
    PLAIN
    -0.07
    hydro
    -0.07
    .parameter
    -0.07
     talk
    -0.07
    नल
    -0.06
    make
    -0.06
    ulated
    -0.06
    POSITIVE LOGITS
     achieves
    0.06
    敢
    0.06
     드
    0.06
     тис
    0.06
     detainees
    0.06
     Pun
    0.06
    願い
    0.06
     Kann
    0.05
    ivate
    0.05
     IDb
    0.05
    Act Density 0.022%

    No Known Activations