© 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. 10-GEMMASCOPE-RES-131K
    7. 22152
    Prev
    Next
    INDEX
    Explanations

    configuration and queries

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

    The neuron fires on code- and markup-style tokens—e.g. programming keywords, file extensions, class or tag names (like “svg,” “exe,” “grep,” “config,” etc.)—i.e. snippets of code or markup.

    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
     refroid
    -2.80
     bascul
    -2.41
     réve
    -2.33
     éta
    -2.30
     coû
    -2.13
     mélang
    -2.09
     régal
    -2.09
     goû
    -2.09
     péné
    -2.08
    쐈
    -2.08
    POSITIVE LOGITS
     and
    4.22
     in
    3.11
     from
    2.94
     by
    2.88
     during
    2.81
     with
    2.81
     including
    2.66
     on
    2.63
     without
    2.63
     all
    2.61
    Activations Density 0.330%

    No Known Activations