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

    Neuronpedia

    INDEX
    Explanations

    try

    np_max-act · gemini-2.0-flash

    This neuron primarily activates on the verb “try” (and its inflected forms like “tries” or “tried”), flagging instances of an attempted action.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Sunset
    -0.06
     حالة
    -0.06
    .Cdecl
    -0.06
    เช
    -0.06
     只
    -0.06
     Values
    -0.06
    obox
    -0.06
    ═
    -0.06
     planetary
    -0.06
    	value
    -0.05
    POSITIVE LOGITS
     trying
    0.10
     attempting
    0.09
     attempted
    0.09
     tries
    0.08
     Trying
    0.08
     attempts
    0.08
    ling
    0.08
    trying
    0.07
    ERING
    0.07
    toc
    0.07
    Act Density 0.039%

    No Known Activations