© 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
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 751
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on numeric‐ or data‐heavy factual statements—i.e. tokens involved in statistics, percentages, years, measurements, and other quantitative information.

    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
     preclude
    0.68
    '
    0.65
     ("
    0.63
     alternately
    0.63
     precludes
    0.63
    --
    0.62
     nominally
    0.62
    _
    0.62
     variously
    0.61
    '?
    0.60
    POSITIVE LOGITS
     Furthermore
    1.85
     Также
    1.75
     Additionally
    1.69
     এছাড়াও
    1.67
     Moreover
    1.66
    Furthermore
    1.61
     Inoltre
    1.59
     또한
    1.53
     However
    1.52
     এছাড়াও
    1.52
    Activations Density 1.399%

    No Known Activations