© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 33293
    Prev
    Next
    INDEX
    Explanations

    code keywords and structure

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

    The neuron fires on programming identifiers—variable and method names (e.g. “command,” “list,” “query,” “getUserAction”).

    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
    loed
    -0.99
     Россий
    -0.98
     Москов
    -0.96
    eventually
    -0.95
    primarily
    -0.93
    mainly
    -0.91
    included
    -0.90
    dians
    -0.90
    erda
    -0.90
     représenter
    -0.89
    POSITIVE LOGITS
     getting
    0.98
     through
    0.96
     chapa
    0.94
    旼
    0.94
     perplexing
    0.93
     within
    0.90
     restrictions
    0.89
     Krieges
    0.89
     choosing
    0.89
     sophisticated
    0.89
    Activations Density 0.071%

    No Known Activations