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

    Neuronpedia

    INDEX
    Explanations

    level

    np_max-act · gemini-2.0-flash

    The neuron detects hyphenated compound terms ending in “-level.”

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Interestingly
    -0.07
     depicted
    -0.06
     Currently
    -0.06
     consisting
    -0.06
    Binary
    -0.06
     hybrid
    -0.06
    енным
    -0.06
    (gcf
    -0.06
    (concat
    -0.06
    ....↵↵
    -0.06
    POSITIVE LOGITS
     step
    0.07
    _encoder
    0.06
     escal
    0.06
     hry
    0.06
    -level
    0.06
     complete
    0.06
     statewide
    0.06
    elems
    0.06
     dign
    0.06
     level
    0.06
    Act Density 0.008%

    No Known Activations