© 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. 13882
    Prev
    Next
    INDEX
    Explanations

    The neuron selectively activates on numeric tokens—especially fragments of floating‐point numbers (decimal values) commonly seen in citations or data.

    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
     ♦
    0.86
    ast
    0.74
    ta
    0.73
     ‚
    0.73
    nath
    0.72
     
    0.72
    ."—
    0.71
    라도
    0.71
    geräte
    0.71
    id
    0.70
    POSITIVE LOGITS
    イナー
    0.78
    詳細は
    0.78
     ప్రకట
    0.77
    スペイン
    0.73
    躾
    0.73
    इन
    0.71
     precisamos
    0.70
     అనంతరం
    0.70
    صل
    0.70
    الحضور
    0.69
    Activations Density 0.001%

    No Known Activations