© 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. 127964
    Prev
    Next
    INDEX
    Explanations

    places and entities

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

    The neuron fires on single‐word nouns that name key entities or fields (e.g. course, database, project, city, organism, protein, METHOD) typically used as table/column identifiers.

    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
     "
    -1.25
     dyr
    -1.14
     atriz
    -1.11
    toch
    -1.11
    diabetes
    -1.11
     papir
    -1.10
    zack
    -1.10
     baby
    -1.09
     professora
    -1.09
     senhora
    -1.09
    POSITIVE LOGITS
     responsabilità
    1.20
    éclairage
    1.16
    indépendance
    1.15
     particolarmente
    1.14
    婀
    1.13
     (
    1.12
    —(
    1.11
     получают
    1.11
     daarbij
    1.09
    enregistrement
    1.09
    Activations Density 0.184%

    No Known Activations