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

    Neuronpedia

    INDEX
    Explanations

    contractions

    np_max-act · gemini-2.0-flash

    the neuron detects common English contractions (words containing an apostrophe like we're, it's, they're, we've).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
     Mobility
    -0.06
     Sky
    -0.06
     Wage
    -0.06
     Cork
    -0.06
     هند
    -0.06
     Fever
    -0.06
     Δι
    -0.06
     Prize
    -0.06
    (headers
    -0.06
     janvier
    -0.06
    POSITIVE LOGITS
    ’m
    0.08
    utory
    0.07
    む
    0.07
    _();↵
    0.07
    eware
    0.07
    'm
    0.07
    Generated
    0.07
     subscribe
    0.07
     handle
    0.06
    成
    0.06
    Act Density 0.248%

    No Known Activations