© 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. 7841
    Prev
    Next
    INDEX
    Explanations

    do not edit by hand

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

    The neuron detects “do not edit by hand”–style warnings in auto-generated code comments.

    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
     accompagner
    -1.72
     bezw
    -1.70
     añadido
    -1.63
    Là
    -1.62
     dát
    -1.61
    쾨
    -1.59
    groet
    -1.56
     arrêter
    -1.55
     všechny
    -1.52
     vão
    -1.52
    POSITIVE LOGITS
     as
    1.85
     although
    1.84
     but
    1.84
     Although
    1.74
     while
    1.67
     in
    1.66
     including
    1.65
    in
    1.58
     However
    1.57
    it
    1.56
    Activations Density 0.029%

    No Known Activations