© 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. 105005
    Prev
    Next
    INDEX
    Explanations

    followed by example descriptions

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

    commas that appear in technical or formal contexts, particularly in patent documents, code, and structured text.

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

    The neuron activates on numeric tokens (numbers, decimal values, and related digit sequences).

    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
     in
    -2.64
     and
    -1.82
     who
    -1.80
    breviations
    -1.63
     would
    -1.59
     before
    -1.58
     on
    -1.58
     not
    -1.52
     to
    -1.51
     because
    -1.49
    POSITIVE LOGITS
     saranno
    1.52
     you
    1.49
     शामिल
    1.46
    ales
    1.44
    明显的
    1.41
     récemment
    1.41
    全家
    1.41
    three
    1.41
    きょう
    1.39
    財布
    1.38
    Activations Density 0.156%

    No Known Activations