© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 9073
    Prev
    Next
    INDEX
    Explanations

    code structure components

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

    The neuron detects identifiers in code that begin with an underscore (e.g. private or “_”-prefixed attributes and methods).

    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
     from
    -2.38
     in
    -2.13
     not
    -2.11
     of
    -2.02
     to
    -1.91
     by
    -1.88
     an
    -1.88
     on
    -1.85
     During
    -1.84
     all
    -1.82
    POSITIVE LOGITS
     alza
    2.34
    伱
    2.28
     maquilla
    2.17
     bascul
    2.05
     corrom
    2.03
    讎
    2.03
    庒
    2.03
     plafon
    2.02
     alarme
    2.02
     zoude
    2.00
    Activations Density 0.004%

    No Known Activations