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

    Neuronpedia

    INDEX
    Explanations

    face

    np_max-act · gemini-2.0-flash

    the neuron activates on words that describe spatial orientation and positioning (e.g. facing, behind, front, back, sideways, angles).

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

    explicit sexual content and descriptions related to intimate situations.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     effort
    -0.07
    les
    -0.07
     vera
    -0.07
     Gio
    -0.06
    вар
    -0.06
     Breath
    -0.06
     yeme
    -0.06
    _mirror
    -0.06
    thesized
    -0.06
     textu
    -0.06
    POSITIVE LOGITS
    face
    0.07
     Francie
    0.06
    -facing
    0.06
     Ames
    0.06
    };↵
    0.06
     forControlEvents
    0.06
    ๕
    0.06
    -Nov
    0.06
     facing
    0.06
    seek
    0.06
    Act Density 0.012%

    No Known Activations