© 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. 54403
    Prev
    Next
    INDEX
    Explanations

    code declarations and variables

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

    The neuron strongly activates on punctuation tokens used for references, dereferences, and grouping in code—e.g. the “&” address-of operator, “*” pointer operator, and parentheses/brackets.

    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
     media
    -0.79
     rb
    -0.79
    alts
    -0.78
    migrationBuilder
    -0.77
     predetermined
    -0.76
     sicherlich
    -0.76
     ジョ
    -0.75
     maps
    -0.75
     storico
    -0.75
     default
    -0.74
    POSITIVE LOGITS
     kdy
    0.84
    leton
    0.82
    störungen
    0.80
     Hell
    0.79
     livre
    0.78
    LETON
    0.77
    ībā
    0.77
     Jordanian
    0.77
     tengo
    0.76
     Центра
    0.76
    Activations Density 0.006%

    No Known Activations