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

    Neuronpedia

    INDEX
    Explanations

    Code and documentation

    np_max-act · gemini-2.0-flash

    The neuron detects code-like syntax tokens and programming-language keywords (i.e., places in the text that look like source code).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
     функци
    -0.07
    τητα
    -0.06
    itant
    -0.06
     conject
    -0.06
    .timeout
    -0.06
     propos
    -0.06
     rehab
    -0.06
    .In
    -0.06
     UTIL
    -0.06
    817
    -0.06
    POSITIVE LOGITS
    ↵
    0.08
    ↵
    0.08
    ")[
    0.08
     ""↵
    0.07
     }])↵
    0.07
    "`↵
    0.07
    0.07
    =''↵
    0.07
    %)↵
    0.07
    "
↵
    0.07
    Act Density 1.147%

    No Known Activations