© 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. 22-GEMMASCOPE-RES-131K
    7. 130807
    Prev
    Next
    INDEX
    Explanations

    This neuron detects LaTeX/math‐mode formatting tokens, in particular backslashes and related command delimiters.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    mathematical integrals and symbols

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    vább
    -1.38
     politiet
    -1.30
     tersebut
    -1.27
     that
    -1.24
    einen
    -1.22
    Trichloroethane
    -1.22
     objectAtIndex
    -1.18
    Lma
    -1.16
    ;%%
    -1.14
     itemprop
    -1.13
    POSITIVE LOGITS
    1.12
     Czechos
    1.03
     frequentemente
    0.99
    ,
    0.99
    </em>
    0.98
    They
    0.98
     they
    0.98
    gese
    0.97
    将来
    0.97
    they
    0.96
    Activations Density 0.006%
    No Known Activations