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

    Neuronpedia

    INDEX
    Explanations

    thing

    np_max-act · gemini-2.0-flash

    This neuron detects the expression “the last thing,” especially when used in phrases like “the last thing to do.”

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     ingin
    -0.06
     dj
    -0.06
     smarter
    -0.06
    	d
    -0.06
    WD
    -0.06
     глуб
    -0.06
     Catalyst
    -0.06
    -base
    -0.06
     slou
    -0.06
     заяв
    -0.06
    POSITIVE LOGITS
    _ASSUME
    0.07
    ریم
    0.07
    _AC
    0.06
    _finished
    0.06
    Finish
    0.06
     Lifestyle
    0.06
     Shaft
    0.06
    _Var
    0.06
     wear
    0.06
    .Min
    0.06
    Act Density 0.011%

    No Known Activations