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

    Neuronpedia

    INDEX
    Explanations

    for

    np_max-act · gemini-2.0-flash

    This neuron detects tokens in self-introductions where the speaker describes themselves as an assistant for a magazine or tabloid (e.g., “my assistant for Cosmopolitan tabloid magazine”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    .Roles
    -0.07
    =result
    -0.06
     regex
    -0.06
    531
    -0.06
     Kur
    -0.06
    _RETURNTRANSFER
    -0.06
     chloride
    -0.06
    .ToLower
    -0.06
     Belarus
    -0.06
     aluminum
    -0.05
    POSITIVE LOGITS
     Securities
    0.07
    emd
    0.07
    иком
    0.06
    編
    0.06
    	echo
    0.06
    нем
    0.06
     欧
    0.06
    _sem
    0.06
    мі
    0.06
     IND
    0.06
    Act Density 0.001%

    No Known Activations