© 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. 22-GEMMASCOPE-RES-131K
    7. 14611
    Prev
    Next
    INDEX
    Explanations

    code structures and technical terms

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

    The neuron strongly responds to programming‐language syntax tokens—especially braces, semicolons, angle brackets, colons, numeric literals, and other punctuation common in 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
     forment
    -1.13
     radikal
    -1.07
     envelo
    -1.06
    Also
    -1.03
     kome
    -1.01
     permanentes
    -0.98
     inoxid
    -0.96
    "/",
    -0.96
     fosfor
    -0.96
     astu
    -0.94
    POSITIVE LOGITS
    さんは
    0.96
     будет
    0.95
     know
    0.95
     synchronization
    0.94
     said
    0.93
     knew
    0.92
     knows
    0.92
     through
    0.87
     portavoz
    0.86
     acknowledges
    0.86
    Activations Density 0.003%

    No Known Activations