© 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. 12433
    Prev
    Next
    INDEX
    Explanations

    economic activities and services

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

    The neuron selectively fires on multi-syllable, domain-specific nouns (often appearing in plural form) that label central scientific or technical concepts.

    oai_token-act-pair · o4-miniTriggered 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
     permitan
    -1.71
    却被
    -1.56
     ausein
    -1.50
    すじ
    -1.49
    ざま
    -1.48
    小孩子
    -1.44
     bandes
    -1.43
    f
    -1.43
     obtenir
    -1.39
    しり
    -1.37
    POSITIVE LOGITS
     about
    1.87
     by
    1.82
     without
    1.80
     Lingkungan
    1.75
     meticulously
    1.73
     Because
    1.71
     freshly
    1.70
     has
    1.69
     does
    1.68
     on
    1.66
    Activations Density 0.134%

    No Known Activations