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

    Neuronpedia

    INDEX
    Explanations

    online forum/technical discussions

    np_max-act · gemini-2.0-flash

    This neuron detects tokens that mark the assistant speaker/role or assistant-turn indicators in a chat transcript.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    LinkId
    -0.07
    	DWORD
    -0.06
    计
    -0.06
     Pussy
    -0.06
     CID
    -0.06
     ethnicity
    -0.06
    	target
    -0.06
     HID
    -0.06
    ouse
    -0.06
    plen
    -0.06
    POSITIVE LOGITS
     invaders
    0.07
     свое
    0.07
     ativ
    0.07
     така
    0.07
     sentient
    0.07
    δρο
    0.06
    自己的
    0.06
    нен
    0.06
    /grpc
    0.06
     luyện
    0.06
    Act Density 0.092%

    No Known Activations