© 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. 22-GEMMASCOPE-RES-131K
    4. 74887
    Prev
    Next
    INDEX
    Explanations

    detail/ or deal in

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

    This neuron fires on single-word lowercase identifiers (nouns) that serve as section or module labels in code—for example “detail,” “security,” “wallet,” “board,” “world,” or “deal.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     péro
    -1.73
     geladeira
    -1.64
    
    -1.59
     cinto
    -1.55
     唯美
    -1.55
     kalem
    -1.55
     suí
    -1.52
     frança
    -1.52
     vijf
    -1.50
     parachoque
    -1.48
    POSITIVE LOGITS
    _
    1.95
     for
    1.88
     as
    1.76
    It
    1.55
    :
    1.52
     at
    1.48
     k
    1.43
    '
    1.43
     give
    1.42
    (
    1.41
    Activations Density 0.001%

    No Known Activations