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

    Neuronpedia

    INDEX
    Explanations

    technical definitions

    np_max-act · gemini-2.0-flash

    This neuron primarily activates on capitalized tokens (proper nouns, acronyms, or other words beginning with an uppercase letter).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    анд
    -0.07
    .original
    -0.07
     congreg
    -0.06
    ROLL
    -0.06
     rectangular
    -0.06
    ौड
    -0.06
    บค
    -0.06
    ीआई
    -0.06
     Zinc
    -0.06
     abc
    -0.06
    POSITIVE LOGITS
    .toJSONString
    0.08
    <object
    0.07
     Threads
    0.07
     Mặt
    0.06
     ]↵↵↵
    0.06
     fixing
    0.06
     Array
    0.06
    AGAIN
    0.06
    .Name
    0.06
    	die
    0.06
    Act Density 0.080%

    No Known Activations