© 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. 10-GEMMASCOPE-RES-131K
    7. 101787
    Prev
    Next
    INDEX
    Explanations

    foreshore, baseman, Hindi party, sphinx, place names

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

    The neuron fires on content‐heavy noun tokens—especially longer or specialized nouns (e.g. party names, position titles, technical terms, Greek‐letter symbols, etc.).

    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
    儅
    -2.78
    鐧
    -2.38
    哢
    -2.30
    ஂ
    -2.28
     sobretudo
    -2.23
    
    -2.20
    嗱
    -2.20
    本章
    -2.20
    mería
    -2.19
    摃
    -2.19
    POSITIVE LOGITS
    The
    2.69
    d
    2.66
    Algunos
    2.58
    _{
    2.39
    E
    2.38
    but
    2.38
     {
    2.33
    a
    2.25
     这个
    2.25
     [
    2.25
    Activations Density 0.023%

    No Known Activations