© 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. 34-GEMMASCOPE-RES-131K
    7. 86763
    Prev
    Next
    INDEX
    Explanations

    meta prefix followed by common suffixes

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

    The neuron detects occurrences of the prefix “meta” (e.g. in words like meta-analysis, MetaMask, metacarpal).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    류
    -0.92
     maybe
    -0.81
    Alfredo
    -0.75
    TextAppearance
    -0.71
     endeav
    -0.71
     supposes
    -0.70
     endeavored
    -0.70
    今は
    -0.69
    今では
    -0.69
    ضب
    -0.69
    POSITIVE LOGITS
    stable
    0.86
    roffenen
    0.84
    stas
    0.82
     Meta
    0.82
    ridescent
    0.79
     оли
    0.79
    Meta
    0.79
    ‌پ
    0.78
    Nanos
    0.78
     oval
    0.77
    Activations Density 0.024%

    No Known Activations