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

    Neuronpedia

    INDEX
    Explanations

    equations

    np_max-act · gemini-2.0-flash

    This neuron detects LaTeX math notation (e.g. commands like \frac, \displaystyle, \mathrm, and other TeX formula tokens).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     د
    -0.07
     regain
    -0.06
     ★
    -0.06
     livest
    -0.06
    erea
    -0.06
    δε
    -0.06
     MU
    -0.06
    entieth
    -0.06
     sure
    -0.06
     belki
    -0.06
    POSITIVE LOGITS
    overnment
    0.07
    Sab
    0.07
    .TXT
    0.06
     extracting
    0.06
    inking
    0.06
    urchases
    0.06
    .ends
    0.06
    uble
    0.06
    opped
    0.06
     Barker
    0.06
    Act Density 0.003%

    No Known Activations