© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 1153
    Prev
    Next
    INDEX
    Explanations

    This neuron specializes in spotting structured metadata entries—labels and their associated numeric codes or dates (e.g., “Category: 1551.5341,” “Action Date: 1392.261,” “Transaction: 615.4069,” barcodes, reference codes, etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ॕ
    0.75
    onneur
    0.68
    endregion
    0.62
     सावधान
    0.61
    irmek
    0.61
    anées
    0.60
    umetric
    0.59
    ികിത്സ
    0.58
    ალური
    0.57
    ຜະລິດຕ
    0.57
    POSITIVE LOGITS
    :
    3.41
    :
    2.38
    :...
    2.23
     :
    2.11
    :[
    2.08
    :{
    2.03
    :.
    1.98
    :}
    1.92
    :,
    1.92
    :'
    1.91
    Activations Density 0.253%

    No Known Activations