© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 628
    Prev
    Next
    INDEX
    Explanations

    elbrus, Chick-fil-A, NetHack

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

    The neuron fires strongly on parts of specific proper nouns and titles (e.g. names of games, bands, people, places, and chemical compounds).

    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_24_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
    <unused75>
    0.27
    ...");
    0.27
     hilfre
    0.27
     nöt
    0.26
     emphatic
    0.26
     dejar
    0.26
     pitiful
    0.26
    mediately
    0.26
     подходя
    0.26
     incremento
    0.26
    POSITIVE LOGITS
     என்னும்
    0.34
    ’
    0.33
    S
    0.31
     adlı
    0.31
     नामक
    0.30
    A
    0.29
    ਸ
    0.29
    と呼ばれる
    0.29
    Ž
    0.29
    Com
    0.29
    Activations Density 0.392%

    No Known Activations