© 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. 12575
    Prev
    Next
    INDEX
    Explanations

    code or programming questions

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

    This neuron fires on long runs of leading whitespace (indentation) in source 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_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
     alho
    -1.02
     炒
    -1.02
    ÁT
    -1.00
     pleaf
    -0.99
     ORIGINAL
    -0.99
    چسب
    -0.99
    protoimpl
    -0.98
     unſ
    -0.98
    蛱
    -0.97
     LUCK
    -0.97
    POSITIVE LOGITS
     what
    1.76
     how
    1.59
     someone
    1.41
     some
    1.40
     there
    1.27
     new
    1.25
     something
    1.22
     this
    1.21
     more
    1.20
     one
    1.20
    Activations Density 0.030%

    No Known Activations