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

    Neuronpedia

    INDEX
    Explanations

    criminal acts and harm

    np_acts-logits-general · gemini-2.5-flash-lite

    references to crime, violence, or weapons, along with related law-enforcement, safety, intent, and legality context.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     বিশ্বের
    0.41
     सुनिश्चित
    0.32
     BeautifulSoup
    0.32
     regex
    0.31
     polyphen
    0.31
     compilers
    0.31
     gracefully
    0.30
     sikk
    0.30
     conquered
    0.30
    中国的
    0.30
    POSITIVE LOGITS
     burglary
    0.50
     robbery
    0.47
     crime
    0.44
     criminal
    0.44
     unlawful
    0.44
     वारदात
    0.43
     felony
    0.43
     homicide
    0.43
     altercation
    0.41
    criminal
    0.41
    Act Density 0.140%

    No Known Activations