© 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. 34-GEMMASCOPE-RES-131K
    7. 64258
    Prev
    Next
    INDEX
    Explanations

    nouns followed by descriptive clauses

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

    The neuron consistently activates on domain-specific content nouns—main subjects or “things” being discussed in the text.

    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
     inſ
    -1.02
    はこんな感じ
    -0.97
    はかなり
    -0.97
     подоб
    -0.95
     aço
    -0.94
    agerie
    -0.92
    再多
    -0.92
     gusta
    -0.91
    anych
    -0.90
    welt
    -0.90
    POSITIVE LOGITS
     that
    4.28
     whose
    2.34
     που
    2.33
     who
    2.30
     with
    2.27
    ที่
    2.14
    ที่มี
    2.13
     deemed
    2.05
     which
    2.02
     که
    1.68
    Activations Density 0.232%

    No Known Activations