© 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-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 67885
    Prev
    Next
    INDEX
    Explanations

    METHOD, SAMPLE, or text following METHOD

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

    This neuron fires on all-caps identifier tokens (e.g. uppercase constant or macro/function names with underscores).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    🧌
    -2.22
    .");
    -2.17
    ☫
    -2.11
    breviations
    -2.08
    鸲
    -2.06
    र
    -2.06
    𝐡
    -2.05
    ৢ
    -2.03
     Presidencia
    -1.99
    䜣
    -1.99
    POSITIVE LOGITS
    <bos>
    2.16
    會
    1.95
    Estamos
    1.88
     him
    1.88
    because
    1.85
    ні
    1.84
     lest
    1.83
    to
    1.80
    what
    1.77
    izing
    1.76
    Activations Density 0.000%

    No Known Activations