© 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. 6925
    Prev
    Next
    INDEX
    Explanations

    dialogue markers and formatting

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

    The neuron spikes on standalone floating‐point numbers or numeric metadata codes (e.g. “2221.1077,” “2021.8931,” etc.) rather than regular 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
    I
    0.63
    ডিয়া
    0.62
    ウォ
    0.61
    Mac
    0.60
    あと
    0.58
    sns
    0.57
     জনের
    0.57
    dragging
    0.57
    A
    0.56
    Col
    0.55
    POSITIVE LOGITS
     ________
    0.60
     ____________
    0.58
     promi
    0.58
     স্বায়
    0.57
     _____________
    0.57
     __________
    0.56
    urgiu
    0.56
    otheby
    0.55
     apparence
    0.54
    _'
    0.54
    Activations Density 0.003%

    No Known Activations