© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 10050
    Prev
    Next
    INDEX
    Explanations

    The neuron selectively responds to numeric tokens—especially digits and time‐stamp or minute markers (e.g. “1,” “min,” “:,” “58th minute”)—commonly found in sports commentary and stat lines.

    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
    دة
    0.99
    тың
    0.96
    ăţ
    0.96
    alarının
    0.96
    دأ
    0.94
    非常的
    0.89
    دت
    0.89
    ových
    0.88
    tim
    0.86
    trap
    0.86
    POSITIVE LOGITS
     specifies
    0.77
     же
    0.74
    二是
    0.73
    ษฐ
    0.72
     clade
    0.72
     вам
    0.70
    で
    0.70
     lɛ
    0.69
     আও
    0.68
     (!)
    0.68
    Activations Density 20.313%

    No Known Activations