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

    Neuronpedia

    INDEX
    Explanations

    code and technical documentation

    np_max-act · gemini-2.0-flash

    The neuron responds to second‐person instructional language—especially “you” plus an action verb (e.g. “You open,” “You go,” “You find”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Hen
    -0.07
     paj
    -0.07
     Sud
    -0.07
    sap
    -0.07
    pent
    -0.07
     ปร
    -0.06
    zek
    -0.06
    旦
    -0.06
     Singapore
    -0.06
    graf
    -0.06
    POSITIVE LOGITS
    errick
    0.07
    ());
↵
    0.07
    detail
    0.07
               ↵
    0.06
    etail
    0.06
    ResultSet
    0.06
                           ↵
    0.06
    .info
    0.06
     Church
    0.06
    255
    0.06
    Act Density 0.000%

    No Known Activations