© 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. 67566
    Prev
    Next
    INDEX
    Explanations

    . followed by word

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

    This neuron is triggered by code‐style punctuation (e.g. the dots in package/import paths, colons, semicolons, etc.), effectively spotting delimiter tokens 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
     Tampoco
    -1.42
    meisters
    -1.34
    erweise
    -1.32
     También
    -1.23
    cracy
    -1.22
     Aún
    -1.21
    kursi
    -1.19
     attraverso
    -1.16
     páss
    -1.16
     terão
    -1.16
    POSITIVE LOGITS
     우리
    1.09
     which
    1.09
     hemel
    1.06
    ฌ
    1.06
     Ленинград
    1.02
    umesc
    1.01
     profesores
    1.01
    %@",
    1.00
    充满了
    0.98
    ˶
    0.98
    Activations Density 0.000%

    No Known Activations