© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 6541
    Prev
    Next
    INDEX
    Explanations

    Laravel and web frameworks

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron strongly fires on long digit sequences and numeric tokens (e.g. version numbers, error‐log codes, cache identifiers, file‐path or URL components) in code or log 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_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
    苺
    0.67
    ElementString
    0.57
    ropa
    0.55
    atania
    0.54
    easter
    0.54
    حدى
    0.52
    IEEE
    0.52
    度が
    0.52
    毕业生
    0.51
    ړو
    0.51
    POSITIVE LOGITS
     artisan
    0.63
     Kazakh
    0.62
     Če
    0.61
     transceiver
    0.61
     প্লা
    0.59
     청
    0.58
     Uzbek
    0.58
    ād
    0.57
     규칙
    0.57
     Artisan
    0.57
    Activations Density 0.011%

    No Known Activations