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

    Neuronpedia

    INDEX
    Explanations

    android

    np_max-act · gemini-2.0-flash

    This neuron recognizes tokens from Android exception stack traces—i.e. package‐qualified class and method names (and related log identifiers) that appear in error dumps.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    -svg
    -0.07
     Indi
    -0.07
     Save
    -0.07
    	object
    -0.07
    Inject
    -0.07
    nat
    -0.07
    _ind
    -0.06
    ortality
    -0.06
    theses
    -0.06
     Java
    -0.06
    POSITIVE LOGITS
     stationed
    0.08
    へと
    0.06
     coordinator
    0.06
     Pioneer
    0.06
    =label
    0.06
     içine
    0.06
    ordinator
    0.06
     hostility
    0.06
    ائمة
    0.06
    에는
    0.06
    Act Density 0.003%

    No Known Activations