© 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. 73127
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions involving square roots

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

    The neuron fires on bits of inline math—variable names, subscripts/superscripts, function names (like exp, Cos), and other LaTeX math tokens.

    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
     to
    -1.10
     which
    -0.95
     elixir
    -0.94
     Aujourd
    -0.93
     Plusieurs
    -0.93
     Hezbollah
    -0.91
     awalnya
    -0.91
    埂
    -0.90
    WindowConstants
    -0.90
    cazione
    -0.90
    POSITIVE LOGITS
     like
    1.18
    dependent
    0.96
    mity
    0.96
    dependencies
    0.94
    like
    0.93
    </b>
    0.92
    apyrus
    0.88
    exchange
    0.86
     tư
    0.86
    empty
    0.85
    Activations Density 0.006%

    No Known Activations