© 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. 129208
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on numeric literals (integers or floating‐point numbers) in code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    code syntax from programming languages

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     aſ
    -0.91
     Tradu
    -0.85
     resuming
    -0.84
    柚子
    -0.83
    捩
    -0.83
    eller
    -0.81
    classList
    -0.77
     раз
    -0.77
    消化
    -0.77
     جمله
    -0.76
    POSITIVE LOGITS
    袄
    1.02
    верное
    0.94
    hews
    0.89
    Läh
    0.88
    δει
    0.88
    verty
    0.85
    cação
    0.85
     uintptr
    0.84
     beiden
    0.82
    isbn
    0.82
    Activations Density 0.021%

    No Known Activations