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

    Neuronpedia

    INDEX
    Explanations

    code/file paths

    np_max-act · gemini-2.0-flash

    The neuron is triggered by subtoken pieces of identifiers containing underscores (e.g. “source_dir”, “dest_dir”) or camel‐style field names (e.g. “shippingAddressFieldGroup”), i.e. it detects variable or field name fragments with underscores or mixed‐case naming patterns.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    .dictionary
    -0.06
    เช
    -0.06
     sweeps
    -0.06
     Fur
    -0.06
    Aliases
    -0.06
     سک
    -0.06
     UIViewController
    -0.06
     Gospel
    -0.06
    .exec
    -0.06
    
    -0.06
    POSITIVE LOGITS
     finden
    0.08
     geçir
    0.07
    	glUniform
    0.07
    ,status
    0.07
     glUniform
    0.06
     examinations
    0.06
     SCIP
    0.06
     sabot
    0.06
     Macedonia
    0.06
     marshal
    0.06
    Act Density 0.106%

    No Known Activations