© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 128628
    Prev
    Next
    INDEX
    Explanations

    code compilation and measurement units

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

    The neuron fires on tokens that are part of file names or file‐extension identifiers (e.g. “.go” or “.txt”) and on command‐invocation keywords associated with those files.

    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
     ajoutez
    -1.84
     éte
    -1.73
     accompagner
    -1.72
    푆
    -1.70
    杙
    -1.70
     soucis
    -1.68
     rafraî
    -1.66
     élar
    -1.63
     Hanya
    -1.62
     corriger
    -1.61
    POSITIVE LOGITS
     and
    4.19
     in
    2.52
     on
    2.41
     or
    2.33
     to
    2.03
     with
    2.00
     including
    1.98
     as
    1.67
     then
    1.67
     at
    1.66
    Activations Density 0.068%

    No Known Activations