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

    Neuronpedia

    INDEX
    Explanations

    subject

    np_max-act · gemini-2.0-flash

    This neuron detects prepositional phrases in procedural descriptions—words like “on” (as in “passed on to”) and “subject” (as in “subject to”) that introduce process steps or conditions.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    увала
    -0.06
     рецеп
    -0.06
    iswa
    -0.06
    _perc
    -0.06
     přech
    -0.06
     ефектив
    -0.06
    ieri
    -0.06
    pile
    -0.06
    MAS
    -0.06
     ráp
    -0.06
    POSITIVE LOGITS
     prone
    0.07
     subject
    0.07
    boundary
    0.07
    렴
    0.07
     subjected
    0.07
     suffers
    0.07
     the
    0.07
     exposure
    0.06
     doivent
    0.06
    omination
    0.06
    Act Density 0.014%

    No Known Activations