© 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. 12831
    Prev
    Next
    INDEX
    Explanations

    specific advice for situations

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

    The neuron fires primarily on tokens that are numeric values (especially floating-point numbers or other multi-digit 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
     কেননা
    0.64
    льзова
    0.60
    казыва
    0.59
    لان
    0.58
    ouvrir
    0.58
    ലിയ
    0.57
     cance
    0.56
    ỏ
    0.56
    парт
    0.55
    isTouching
    0.55
    POSITIVE LOGITS
     ឬ
    0.59
    facts
    0.57
    的其他
    0.56
    彻
    0.55
     ಅಥವಾ
    0.55
     facts
    0.54
     அல்லது
    0.54
     чисел
    0.54
     পরিস্থিতি
    0.53
     nebo
    0.53
    Activations Density 0.001%

    No Known Activations