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

    Neuronpedia

    INDEX
    Explanations

    auxiliary verbs

    np_max-act · gemini-2.0-flash

    The neuron detects words and phrases used in the assistant’s disclaimers about inability or impossibility (e.g. “not,” “impossible,” “can’t,” “something that can’t be done overnight,” etc.).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    LOY
    -0.07
    ,this
    -0.06
     refine
    -0.06
     distributions
    -0.06
     Ston
    -0.06
     insurers
    -0.06
    =req
    -0.06
     sources
    -0.06
     nostra
    -0.06
    Notifications
    -0.06
    POSITIVE LOGITS
     sidelined
    0.07
    _part
    0.07
     cunt
    0.07
     οπο
    0.06
     getEmail
    0.06
    uenta
    0.06
    vious
    0.06
    _STS
    0.06
     usize
    0.06
    ideo
    0.06
    Act Density 0.020%

    No Known Activations