© 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. 39505
    Prev
    Next
    INDEX
    Explanations

    ordering words/numbers

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

    The neuron is attuned to document‐structure markers—particularly tokens from section titles, headings, and labeled identifiers.

    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
     Dinas
    -1.24
     sogenannten
    -1.22
     attravers
    -1.18
     alcune
    -1.17
     scorso
    -1.16
     alguns
    -1.15
     bedien
    -1.14
     insanely
    -1.13
     insanların
    -1.12
    
    -1.07
    POSITIVE LOGITS
     first
    1.51
     primera
    1.39
     now
    1.31
     even
    1.30
     new
    1.27
     потім
    1.25
    和
    1.23
     després
    1.21
     their
    1.20
     our
    1.20
    Activations Density 0.011%

    No Known Activations