© 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. 22-GEMMASCOPE-RES-131K
    7. 95747
    Prev
    Next
    INDEX
    Explanations

    places and amenities

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

    The neuron fires most strongly on countable concrete nouns naming areas, venues, or categories (e.g. “area,” “shops,” “species,” “presidents,” “educators,” etc.).

    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
     if
    -1.80
     in
    -1.73
     two
    -1.67
     slightly
    -1.59
     by
    -1.50
     or
    -1.48
     it
    -1.47
     shortly
    -1.46
     (
    -1.46
     possibly
    -1.45
    POSITIVE LOGITS
     diverses
    1.57
    hundreds
    1.55
     ſur
    1.45
     maniere
    1.41
    thousands
    1.41
     rése
    1.41
     jezt
    1.38
    ⸱
    1.36
     素描
    1.36
     niedersachsen
    1.35
    Activations Density 0.064%

    No Known Activations