© 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. 112640
    Prev
    Next
    INDEX
    Explanations

    programming code

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

    the word "code" in programming contexts, particularly in variable names, JSON keys, and code snippets.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    This neuron picks out numeric literals (integer or floating‐point numbers) in code or technical text.

    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
    
    -1.43
    艄
    -1.37
    dónde
    -1.35
     acheter
    -1.33
    翃
    -1.31
    を中心
    -1.30
    方才
    -1.30
     suche
    -1.29
     basit
    -1.28
    manera
    -1.28
    POSITIVE LOGITS
    You
    2.41
    There
    2.14
    Some
    2.09
    They
    2.06
    It
    2.02
    Not
    1.84
    With
    1.83
    What
    1.80
    Maybe
    1.77
    Being
    1.76
    Activations Density 0.013%

    No Known Activations