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

    Neuronpedia

    INDEX
    Explanations

    varied types of writing

    np_max-act · gemini-2.0-flash

    The neuron flags the part of a user’s instruction where they define the target subject or persona for the requested content (i.e. the clause that specifies who or what the output should be about).

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

    documents related to recommendations and endorsements.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    style
    -0.06
     Year
    -0.06
    igue
    -0.06
     vita
    -0.06
    /train
    -0.06
     Hed
    -0.06
     �
    -0.06
    (p
    -0.06
     year
    -0.06
    .create
    -0.06
    POSITIVE LOGITS
    ใด
    0.07
    EMPLARY
    0.07
     způsobem
    0.07
     Adresse
    0.07
    _TIMES
    0.06
     jednoduch
    0.06
     verst
    0.06
     FormGroup
    0.06
    modification
    0.06
    erken
    0.06
    Act Density 0.121%

    No Known Activations