© 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. 22-GEMMASCOPE-RES-131K
    7. 65186
    Prev
    Next
    INDEX
    Explanations

    code instantiation with parentheses

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

    The neuron fires on numeric literals (numbers) appearing in code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     and
    -1.11
    رت
    -0.99
     the
    -0.99
    لل
    -0.98
    holde
    -0.98
    efois
    -0.97
     or
    -0.97
    holms
    -0.94
    雪
    -0.94
     ekonomi
    -0.92
    POSITIVE LOGITS
    いましたが
    1.16
    的吗
    1.07
    などは
    1.02
    があって
    1.00
     necessários
    0.97
    ローブ
    0.97
     accueillir
    0.94
    )))));
    0.94
    を作り
    0.93
     naudoti
    0.92
    Activations Density 0.046%

    No Known Activations