© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 105273
    Prev
    Next
    INDEX
    Explanations

    code syntax delimiters

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

    generic type declarations in programming code, particularly class or function definitions with type parameters enclosed in angle brackets.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron fires on class‐declaration lines that specify inheritance or implementation (e.g. “extends” or “implements” clauses).

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

    code declarations—especially class/interface or function signatures involving inheritance/generics—and the opening brace that begins their bodies.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     globales
    -1.77
     mentales
    -1.72
    6
    -1.71
    3
    -1.61
    鵙
    -1.60
    "]();
    -1.57
     traduc
    -1.55
    нование
    -1.55
    1
    -1.53
    7
    -1.50
    POSITIVE LOGITS
     that
    2.38
     kunnen
    1.80
     or
    1.78
    :
    1.77
    ;
    1.67
    {
    1.66
    tratamiento
    1.55
    …
    1.55
     to
    1.54
    —
    1.54
    Activations Density 0.016%

    No Known Activations