© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 35982
    Prev
    Next
    INDEX
    Explanations

    design names, prime numbers, unisex names, school types

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

    The neuron is strongly activated by the substring “edu” (e.g. in “.edu” domain names).

    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
     out
    -1.35
     over
    -1.27
     one
    -1.27
     which
    -1.17
     use
    -1.16
     procedures
    -1.14
     both
    -1.14
     most
    -1.13
     their
    -1.12
     as
    -1.11
    POSITIVE LOGITS
     alcune
    1.31
     alcuni
    1.23
    মা
    1.21
     magnific
    1.20
    aien
    1.16
     nombreux
    1.09
     tumultuous
    1.09
     opulent
    1.08
    vät
    1.08
     jego
    1.07
    Activations Density 0.017%

    No Known Activations