© 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. 44827
    Prev
    Next
    INDEX
    Explanations

    England and Wales, or gynecology

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

    names of regions, territories, or locations, particularly focusing on states, countries, and specialized fields.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron activates on technical or organizational proper nouns and acronyms—typically capitalized multi-token terms (often with numbers or specialized suffixes) such as “CRISPR/Cas9,” journal or department names, and other formal titles.

    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
     or
    -2.13
     without
    -1.76
     new
    -1.73
     at
    -1.70
     and
    -1.70
    ......
    -1.69
     with
    -1.69
    ........
    -1.65
     because
    -1.64
     gets
    -1.63
    POSITIVE LOGITS
     appeler
    1.48
     plafon
    1.46
     ajouter
    1.39
    絝
    1.38
     créer
    1.38
     salari
    1.37
     faciliter
    1.36
    ‘’
    1.36
     AMAZING
    1.34
     يمكنك
    1.34
    Activations Density 0.112%

    No Known Activations