© 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. 34-GEMMASCOPE-RES-131K
    7. 90398
    Prev
    Next
    INDEX
    Explanations

    # include directives

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

    The neuron strongly activates on code‐style tokens—especially underscores, version numbers, and other programming/metadata keywords (e.g. “version,” “_,” “#include”)—i.e. it identifies snippets of source 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_34/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
     Bayes
    -0.71
    大于
    -0.71
     Conclusions
    -0.71
    小于
    -0.71
     ngào
    -0.70
    DAYS
    -0.70
    وسف
    -0.69
    Eksterne
    -0.69
    nombre
    -0.68
     Frances
    -0.68
    POSITIVE LOGITS
     conditionally
    0.72
     zmian
    0.69
     trzy
    0.66
     disinterested
    0.65
     Porto
    0.65
     psychiatric
    0.65
     грузо
    0.63
     exoner
    0.63
    Indexer
    0.62
     کسی
    0.61
    Activations Density 1.930%

    No Known Activations