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

    Neuronpedia

    INDEX
    Explanations

    Code/terminal commands

    np_max-act · gemini-2.0-flash

    This neuron fires on tokens that are names of code-level commands or functions (e.g. assertion calls like “assert_ok,” wrapper scripts, or other invoked function/command identifiers).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     ファ
    -0.07
    е�
    -0.07
    sword
    -0.06
    important
    -0.06
    PLL
    -0.06
    arding
    -0.06
    UserService
    -0.06
    've
    -0.06
    .currentTime
    -0.06
     strapped
    -0.06
    POSITIVE LOGITS
     possibly
    0.07
     Mutation
    0.07
     etkin
    0.07
     quir
    0.06
     novelty
    0.06
     hafif
    0.06
     informational
    0.06
     dijital
    0.06
    Gravity
    0.06
     leven
    0.06
    Act Density 0.019%

    No Known Activations