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

    Neuronpedia

    INDEX
    Explanations

    must

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on the auxiliary “will,” marking future‐tense/modal statements.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Pett
    -0.07
     GameManager
    -0.07
    Clear
    -0.07
     Bent
    -0.07
     Planner
    -0.07
     Terminator
    -0.06
    ButtonText
    -0.06
    _shot
    -0.06
     P
    -0.06
    -0.06
    POSITIVE LOGITS
     writers
    0.06
     writer
    0.06
     rog
    0.06
     thù
    0.06
     hel
    0.06
     πριν
    0.06
     eigentlich
    0.06
    {-#
    0.06
     Mouth
    0.06
    rophe
    0.06
    Act Density 0.022%

    No Known Activations