© 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
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 15897
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on polite request or question phrases (e.g. “Could you…”, “Can I…”, “Also, could those be…?”).

    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
    を楽しむ
    0.64
     geçmiş
    0.52
     смотря
    0.52
    ƌ
    0.52
     competition
    0.52
     displacements
    0.52
    謨
    0.52
     undergoes
    0.51
     ഹോളി
    0.51
     تحصل
    0.50
    POSITIVE LOGITS
     please
    4.19
    please
    3.95
     PLEASE
    3.73
     Please
    3.66
    Please
    3.63
    PLEASE
    3.39
     пожалуйста
    3.22
    请
    3.22
     plz
    3.11
    請
    2.97
    Activations Density 0.393%

    No Known Activations