© 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. 10-GEMMASCOPE-RES-131K
    7. 99575
    Prev
    Next
    INDEX
    Explanations

    mathematical symbols and structures

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

    The neuron fires on LaTeX font‐style macro names (e.g. “cal,” “mathfrak,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    1
    -2.45
     ولكن
    -2.20
    勍
    -2.14
     envía
    -1.92
    詒
    -1.79
     אבל
    -1.78
    meist
    -1.73
    ᷆
    -1.73
     но
    -1.70
     pero
    -1.70
    POSITIVE LOGITS
    {
    2.92
    }
    2.22
     we
    2.08
    出来た
    2.02
     the
    2.02
    各样的
    1.91
     estekak
    1.83
    1.72
    顷
    1.71
    }^\
    1.70
    Activations Density 0.002%

    No Known Activations