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

    Neuronpedia

    INDEX
    Explanations

    F

    np_max-act · gemini-2.0-flash

    This neuron activates on the all-caps courtroom header tokens (abbreviations like AP, F, C in the “IN THE UNITED STATES COURT OF APPEALS FOR THE FIFTH CIRCUIT” heading).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    boilerplate header and metadata from U.S. Court of Appeals opinions—especially Fifth Circuit—such as court/circuit names, case numbers, filing dates, and procedural headings.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     naš
    -0.07
    언
    -0.07
     روابط
    -0.07
    альних
    -0.06
     nữa
    -0.06
     düşür
    -0.06
     सन
    -0.06
     trung
    -0.06
     Pett
    -0.06
     روح
    -0.06
    POSITIVE LOGITS
     mortgage
    0.07
    _payload
    0.07
    -tested
    0.07
     adverts
    0.07
    Conflict
    0.07
     tempted
    0.06
    δή
    0.06
     graduates
    0.06
     відбу
    0.06
     nær
    0.06
    Act Density 0.007%

    No Known Activations