© 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. 34-GEMMASCOPE-RES-131K
    7. 77838
    Prev
    Next
    INDEX
    Explanations

    Python code imports

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

    This neuron selectively activates on programming‐code tokens—especially common identifiers and keywords (like import, init, from, in, paths, etc.)—marking “code” elements rather than ordinary words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     венти
    -0.82
     conditioners
    -0.80
    jarah
    -0.79
    LEI
    -0.79
     voork
    -0.77
     Architektur
    -0.76
     Polytechnic
    -0.75
     Krise
    -0.74
     caja
    -0.72
     văn
    -0.72
    POSITIVE LOGITS
    ség
    0.95
    scss
    0.93
    </s>
    0.89
    vosti
    0.88
     Westen
    0.88
    ,_
    0.88
    Один
    0.83
    Эта
    0.82
     amerik
    0.82
    tamment
    0.82
    Activations Density 0.050%

    No Known Activations