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

    Neuronpedia

    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    The neuron is triggered by formatting tokens—especially list‐item markers (dashes), numbered prefixes, and other punctuation used to structure instructions.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Stats
    -0.06
    ,E
    -0.06
    ып
    -0.06
     wildcard
    -0.06
    iphertext
    -0.06
     Contracts
    -0.06
    HeadersHeight
    -0.06
     pelic
    -0.06
     oben
    -0.06
     Ingram
    -0.06
    POSITIVE LOGITS
    -native
    0.07
     ");↵
    0.07
    Prosec
    0.06
     Kab
    0.06
     เมตร
    0.06
    	TArray
    0.06
     Fate
    0.06
     Performing
    0.06
    ');");↵
    0.06
    RTOS
    0.06
    Act Density 0.020%

    No Known Activations