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

    Neuronpedia

    INDEX
    Explanations

    Explanations

    np_max-act · gemini-2.0-flash

    This neuron activates on explanatory, pedagogical language—phrases that describe helping or showing how something works (e.g. “help us understand how,” “looks at how,” “can help us understand”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    إن
    -0.07
     discs
    -0.07
     awake
    -0.07
    Mana
    -0.06
    bg
    -0.06
    배
    -0.06
     Broad
    -0.06
     aside
    -0.06
     denote
    -0.06
     Atmos
    -0.06
    POSITIVE LOGITS
    cdot
    0.06
    "],["
    0.06
    okableCall
    0.06
     owner
    0.06
     سود
    0.06
    	ast
    0.06
    \Carbon
    0.06
    -symbol
    0.05
    _CODE
    0.05
    ième
    0.05
    Act Density 0.192%

    No Known Activations