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

    Neuronpedia

    INDEX
    Explanations

    Riots and clashes

    np_max-act · gemini-2.0-flash

    This neuron detects occurrences of the word “blasphemy” (and its sub‐token parts) in the text.

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

    references to communal violence or riots, particularly involving Hindus and Muslims.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     الثاني
    -0.07
     Міністер
    -0.07
    itivity
    -0.06
    ++]=
    -0.06
    	status
    -0.06
     хочу
    -0.06
    ｯ
    -0.06
    _indicator
    -0.06
     :)↵↵
    -0.06
     dozens
    -0.06
    POSITIVE LOGITS
     catastrophe
    0.07
    čná
    0.07
    фици
    0.06
    .Once
    0.06
     Cannabis
    0.06
    igrated
    0.06
     Nhật
    0.06
     Comey
    0.06
     projektu
    0.06
    .addAttribute
    0.06
    Act Density 0.033%

    No Known Activations