© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 4352
    Prev
    Next
    INDEX
    Explanations

    mathematical terms or symbols

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

    The neuron responds to LaTeX/math operator tokens (e.g. “dist”, “Mat”, “Coker”)—i.e. it detects pieces of math notation.

    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
    Archite
    -0.88
     insiste
    -0.87
     ensued
    -0.85
    obses
    -0.83
    結婚
    -0.83
     diğer
    -0.83
    Abdul
    -0.81
    experiences
    -0.80
    Nada
    -0.80
    Nea
    -0.80
    POSITIVE LOGITS
    ⁡
    1.12
    after
    1.03
     $_{\
    1.00
     $_{
    0.98
    r
    0.92
    all
    0.90
    n
    0.89
     why
    0.89
    new
    0.88
    l
    0.87
    Activations Density 0.023%

    No Known Activations