© 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. 85305
    Prev
    Next
    INDEX
    Explanations

    code keywords and syntax

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

    The neuron strongly activates on very short, terse identifier names in code—especially common abbreviated variable names like ptr, buf, idx, x, pair, etc.

    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
     =
    -2.02
     in
    -1.62
     on
    -1.48
    犼
    -1.45
     entreprises
    -1.41
     شنبه
    -1.41
    镡
    -1.39
     знать
    -1.35
     geel
    -1.31
     today
    -1.30
    POSITIVE LOGITS
    だそうです
    1.44
    を注文
    1.41
     Some
    1.38
     Moreover
    1.38
    They
    1.33
     огром
    1.31
     المختلفة
    1.30
     podremos
    1.28
    Moreover
    1.27
    which
    1.25
    Activations Density 0.069%

    No Known Activations