© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 14103
    Prev
    Next
    INDEX
    Explanations

    code syntax and programming

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

    The neuron fires on code operator punctuation—things like the “.” for member access, “[" for indexing, “=” and “+” for assignment/arithmetic—i.e. syntactic operator tokens.

    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
     Dinas
    -1.09
     وكانت
    -1.08
    撂
    -1.02
    ных
    -1.00
     SOLAR
    -1.00
     Womens
    -1.00
    
    -1.00
    骉
    -0.99
    ཥ
    -0.98
    厛
    -0.96
    POSITIVE LOGITS
     these
    1.46
     those
    1.20
     some
    1.16
     this
    1.12
     new
    1.06
     on
    1.06
     people
    1.02
     old
    0.97
     There
    0.93
    there
    0.93
    Activations Density 0.002%

    No Known Activations