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

    Neuronpedia

    INDEX
    Explanations

    negative/critical sentiment

    np_max-act · gemini-2.0-flash

    The neuron fires on bluntly negative or insulting terms—especially profanity and pejoratives (e.g. “bullshit,” “brainwashed,” “dogma,” “subhuman”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     molecular
    -0.07
     Webb
    -0.07
     béné
    -0.06
    truth
    -0.06
    orce
    -0.06
     Ill
    -0.06
     moons
    -0.06
     Jain
    -0.06
     rally
    -0.06
     Validation
    -0.06
    POSITIVE LOGITS
     continuation
    0.07
     peuvent
    0.06
    metis
    0.06
     nomin
    0.06
     extremist
    0.06
     ул
    0.06
    \htdocs
    0.06
     vt
    0.06
    bits
    0.06
    bios
    0.06
    Act Density 0.132%

    No Known Activations