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

    Neuronpedia

    INDEX
    Explanations

    ingredient

    np_max-act · gemini-2.0-flash

    The neuron is selectively activating on the section-header token “Ingredient(s),” marking the start of an ingredients list in a recipe.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    وذ
    -0.08
     Assessment
    -0.07
     kiểm
    -0.06
    codes
    -0.06
     exchanges
    -0.06
     seal
    -0.06
    Paul
    -0.06
     test
    -0.06
    adoras
    -0.06
     RSA
    -0.06
    POSITIVE LOGITS
     ingredients
    0.13
     ingredient
    0.12
     Ingredient
    0.09
    Credentials
    0.08
    .inst
    0.08
    (credentials
    0.07
     �
    0.07
     Ingredients
    0.07
    ingredient
    0.07
    redentials
    0.07
    Act Density 0.007%

    No Known Activations