© 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. 65213
    Prev
    Next
    INDEX
    Explanations

    mathematical variables and symbols

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

    The neuron fires on tokens that are part of mathematical formulae—especially LaTeX-style math symbols, variable names (like f or v), subscripts (the underscore), and keywords such as “forall.”

    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
     esimerkiksi
    -1.71
     honlap
    -1.53
     coût
    -1.52
    탉
    -1.51
     lokale
    -1.51
     veulent
    -1.49
    irée
    -1.49
     refroid
    -1.49
     voient
    -1.49
     traité
    -1.48
    POSITIVE LOGITS
     for
    2.23
     which
    2.03
     to
    2.00
     being
    1.81
     It
    1.63
    },
    1.63
     and
    1.62
     The
    1.60
    We
    1.60
    甭
    1.59
    Activations Density 0.061%

    No Known Activations