© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 14148
    Prev
    Next
    INDEX
    Explanations

    French, Spanish, and English introductory phrases

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

    This neuron responds to non‐regular lowercase words—especially tokens that start with an uppercase letter, digits, or punctuation (e.g. proper names, acronyms, numbers, slashes).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     incrí
    -1.40
     réfrig
    -1.35
     Logement
    -1.35
     célè
    -1.33
     pistolet
    -1.32
    鮰
    -1.32
     anadolu
    -1.31
    untuk
    -1.28
     malé
    -1.27
     faciliter
    -1.27
    POSITIVE LOGITS
     with
    1.84
     will
    1.57
     in
    1.50
     without
    1.50
     as
    1.48
     does
    1.36
     has
    1.36
     should
    1.36
     when
    1.36
     there
    1.34
    Activations Density 0.000%

    No Known Activations