© 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. 31-GEMMASCOPE-2-RES-16K
    7. 8141
    Prev
    Next
    INDEX
    Explanations

    The neuron selectively activates on numeric tokens (years, counts, or other numerals).

    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
    ৎস
    0.65
     ワイド
    0.60
    繍
    0.59
    າຍ
    0.58
     পাগ
    0.58
    ві
    0.57
    终止
    0.57
     cierre
    0.57
     Kxh
    0.56
     ケ
    0.56
    POSITIVE LOGITS
     debut
    1.05
     debuts
    0.96
     debuted
    0.96
     debuting
    0.95
     Debut
    0.91
    debut
    0.88
    デビュー
    0.87
    Debut
    0.85
     деб
    0.75
     första
    0.75
    Activations Density 0.317%

    No Known Activations