© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 130813
    Prev
    Next
    INDEX
    Explanations

    programming code comments or identifiers

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

    The neuron detects programming constructs that use angle-bracket syntax to specify generic or template types (for example, “function<…>”).

    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
     golongan
    -1.60
     pejabat
    -1.55
     its
    -1.54
     pelajar
    -1.49
     you
    -1.48
    ut
    -1.40
     byste
    -1.40
     pegawai
    -1.38
    最重要的
    -1.37
     brachten
    -1.37
    POSITIVE LOGITS
    </strong>
    1.66
    .
    1.57
    人间
    1.48
    Tuy
    1.45
    也
    1.35
    Iya
    1.32
    Akan
    1.25
     szkole
    1.23
     höhere
    1.22
     bosco
    1.22
    Activations Density 0.000%

    No Known Activations