© 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. 22-GEMMASCOPE-RES-131K
    7. 128058
    Prev
    Next
    INDEX
    Explanations

    The neuron selectively activates on semantically rich, often multisyllabic content words—especially abstract nouns, verbs, and gerunds conveying thematic or conceptual meaning.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    states of change and time

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    亵
    -1.27
    ό
    -1.23
    跻
    -1.23
     secrétaire
    -1.20
    </h3>
    -1.20
    クロー
    -1.20
    depuis
    -1.19
     стомато
    -1.17
    ányi
    -1.16
    ֶּ
    -1.15
    POSITIVE LOGITS
     something
    1.38
     bílé
    1.36
     baweł
    1.20
    ちょっと
    1.16
    Etimología
    1.14
     oneself
    1.14
     suplementos
    1.14
     tanques
    1.12
     people
    1.10
     criaturas
    1.10
    Activations Density 0.038%

    No Known Activations