© 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. 10129
    Prev
    Next
    INDEX
    Explanations

    let me know

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

    The neuron detects direct invitation questions or calls to action phrased like “Do you want to …?” or “Want to …?”

    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
     querer
    0.87
     WANT
    0.82
     нужно
    0.77
    NEEDED
    0.77
     vouloir
    0.74
    要去
    0.72
     Нужно
    0.72
     WANTED
    0.71
    想要的
    0.71
     хочу
    0.69
    POSITIVE LOGITS
     know
    0.72
    ა
    0.69
    知道
    0.66
    ทราบ
    0.64
     hear
    0.64
     ensure
    0.63
     знать
    0.62
     maximize
    0.61
    特别是
    0.61
     minimize
    0.60
    Activations Density 0.190%

    No Known Activations