© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 66230
    Prev
    Next
    INDEX
    Explanations

    code structure and identifiers

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

    The neuron fires primarily on numeric literal tokens (e.g. integer or floating‐point constants) in the code.

    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
     or
    -2.06
     После
    -1.73
     червня
    -1.63
    
    -1.63
     вересня
    -1.54
     梁
    -1.53
     and
    -1.52
     октября
    -1.52
     continuously
    -1.51
     from
    -1.50
    POSITIVE LOGITS
     he
    1.88
    訁
    1.73
     who
    1.66
    to
    1.62
     
    1.59
    {
    1.58
    (
    1.52
     schein
    1.52
     intention
    1.48
    秶
    1.48
    Activations Density 0.017%

    No Known Activations