© 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. 34-GEMMASCOPE-RES-131K
    7. 59267
    Prev
    Next
    INDEX
    Explanations

    nationalities, religions, and political affiliations

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

    The neuron strongly activates on capitalized content words—proper nouns, titles, and group or identity labels.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    人たち
    -1.20
     competente
    -1.08
     makers
    -1.08
     monstrous
    -1.02
    凄く
    -1.02
     ſmall
    -1.02
    ݀
    -1.01
    聴いて
    -1.00
     sindic
    -0.98
     coté
    -0.97
    POSITIVE LOGITS
     who
    2.53
     whose
    1.34
     with
    1.30
     whom
    1.11
    who
    1.10
     in
    1.04
     христи
    1.04
    strator
    0.99
     from
    0.96
     ollut
    0.93
    Activations Density 0.124%

    No Known Activations