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

    Neuronpedia

    INDEX
    Explanations

    Modality/auxiliary verbs

    np_max-act · gemini-2.0-flash

    This neuron detects auxiliary-verb constructions (e.g., “is improved,” “will take”)—verbs preceded by helping verbs that signal processes or changes.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    (identity
    -0.07
    ningar
    -0.06
    ermal
    -0.06
    that
    -0.06
    >x
    -0.06
     зрост
    -0.06
    culos
    -0.06
     tangible
    -0.06
    /books
    -0.06
     Such
    -0.06
    POSITIVE LOGITS
     hoàn
    0.07
    opard
    0.07
     endoth
    0.06
    ??↵↵
    0.06
    -admin
    0.06
     Grip
    0.06
    まと
    0.06
    ikon
    0.06
    南省
    0.06
     Extension
    0.06
    Act Density 0.142%

    No Known Activations