© 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. 130994
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on Swift code declarations—especially type and generic declarations (class, struct, enum, protocol, func with generic `<T>` and `where` clauses).

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

    places and names, scientific and code-related topics

    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
     japonaise
    -1.05
     prédé
    -1.02
    法國
    -1.02
     exclusivement
    -1.01
    mô
    -0.98
    contoh
    -0.98
     Bibliote
    -0.97
    ॑
    -0.96
    Üdv
    -0.96
    şu
    -0.96
    POSITIVE LOGITS
    w
    1.13
    on
    1.08
     lleva
    1.07
     режима
    1.07
    inerary
    1.06
    can
    1.06
     who
    1.05
    ophobic
    1.05
    en
    1.04
    arsal
    1.04
    Activations Density 0.004%

    No Known Activations