© 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. 26183
    Prev
    Next
    INDEX
    Explanations

    variable declaration keywords

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

    The neuron is specialized to spot function‐call syntax (identifiers immediately followed by an opening parenthesis) 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
     lutar
    -1.04
    ologische
    -0.84
    žil
    -0.82
     نهر
    -0.81
    financi
    -0.81
     zám
    -0.79
     piemē
    -0.77
    INCLUD
    -0.77
    lover
    -0.77
     primero
    -0.77
    POSITIVE LOGITS
     rire
    0.94
    に対応
    0.93
    ълга
    0.90
    ɚ
    0.88
    *}[
    0.88
     DialogInterface
    0.87
     names
    0.87
     that
    0.86
    nombres
    0.85
     name
    0.84
    Activations Density 0.003%

    No Known Activations