© 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. 22-GEMMASCOPE-RES-131K
    7. 39430
    Prev
    Next
    INDEX
    Explanations

    introduction

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

    The neuron is specialized to spot structural markers and section‐heading tokens (e.g. the “---” and “***” separators, keyword/abstract headings, and other formatting delimiters) in scholarly documents.

    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
     then
    -1.18
    ,
    -1.00
    DONALD
    -0.98
    on
    -0.98
    <i>
    -0.95
     }));
    -0.92
    OGA
    -0.91
     biograf
    -0.90
    <em>
    -0.88
     siècles
    -0.88
    POSITIVE LOGITS
     utilizzo
    1.16
     lanzó
    1.11
     enviado
    0.97
    Kval
    0.95
     publicó
    0.95
     volvió
    0.89
    malıdır
    0.89
     debole
    0.89
     manej
    0.88
    Throwaway
    0.87
    Activations Density 0.046%

    No Known Activations