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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 225
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on standalone numeral tokens (e.g. “1,” “2,” etc.), i.e. it detects list or chapter/item numbers in the text.

    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_24_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
    ৯
    0.47
    ৪
    0.41
    ש
    0.40
    너무
    0.40
    ९
    0.39
    ppled
    0.39
    ו
    0.38
    ง
    0.38
    zelfde
    0.38
    quele
    0.38
    POSITIVE LOGITS
    st
    0.52
    shire
    0.38
    ship
    0.38
    ación
    0.36
    り
    0.35
    markt
    0.34
    ্লাহ
    0.34
     దీంతో
    0.33
    ness
    0.32
    Alchemy
    0.32
    Activations Density 0.000%

    No Known Activations