© 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. 24-GEMMASCOPE-2-RES-16K
    4. 7610
    Prev
    Next
    INDEX
    Explanations

    d<number>

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

    The neuron strongly activates on very high‐frequency function words (common stopwords like “and,” “a,” “on,” possessive ’s, etc.).

    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
     chhoti
    0.98
     Еще
    0.97
     charmed
    0.77
     Seh
    0.77
     باقي
    0.77
    횐
    0.77
    माइंडर
    0.76
    薁
    0.75
    gqlgen
    0.75
     Mongolia
    0.75
    POSITIVE LOGITS
    ism
    0.84
     transcendence
    0.79
    кі
    0.79
    att
    0.73
    က
    0.73
    ard
    0.73
    ви
    0.70
     piecewise
    0.69
    </h3>
    0.68
    淒
    0.68
    Activations Density 0.001%

    No Known Activations