© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 112447
    Prev
    Next
    INDEX
    Explanations

    academic or informational context retrieval

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

    formatting elements in structured documents, particularly punctuation marks, numbers, and symbols that appear in academic or technical text layouts (like colons, line breaks, dates, and section markers).

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

    The neuron activates strongly on numeric tokens (digits, years, or other numbers) in the text.

    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
     sabun
    -1.81
     menyen
    -1.80
    𝕿
    -1.77
     Canadá
    -1.76
     miliki
    -1.74
     novelas
    -1.70
     menik
    -1.70
     bendera
    -1.70
     Japón
    -1.69
     februar
    -1.69
    POSITIVE LOGITS
     
    1.48
     years
    1.38
    ↵
    1.38
    –
    1.37
    <b>
    1.35
    <strong>
    1.34
     later
    1.34
     already
    1.28
     study
    1.28
     involvement
    1.27
    Activations Density 0.525%

    No Known Activations