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

    Neuronpedia

    INDEX
    Explanations

    Filling tightly

    np_max-act · gemini-2.0-flash

    the neuron activates on words that describe densely packing or stuffing content (e.g. “cram,” “crammed”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    inction
    -0.07
    GLOSS
    -0.07
    πι
    -0.07
     zombies
    -0.07
     distinction
    -0.06
    abile
    -0.06
    Song
    -0.06
    Previous
    -0.06
    Calls
    -0.06
    umbs
    -0.06
    POSITIVE LOGITS
     tightened
    0.07
    ovně
    0.06
    凝
    0.06
     Г
    0.06
     Det
    0.06
    �
    0.06
     ini
    0.06
     �
    0.06
    steder
    0.06
     Imp
    0.06
    Act Density 0.009%

    No Known Activations