© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 79769
    Prev
    Next
    INDEX
    Explanations

    starting or beginning actions

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

    The neuron preferentially activates on methodological or procedural language—terms that describe techniques, experiments, or steps in a technical/scientific process.

    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
     mismas
    -1.55
     getan
    -1.36
     sauvages
    -1.25
     vertes
    -1.23
     wody
    -1.16
    拶
    -1.16
     Variante
    -1.13
     decisiones
    -1.13
     conclusione
    -1.13
     lorsqu
    -1.13
    POSITIVE LOGITS
    migrationBuilder
    1.28
     Starting
    1.20
    ólicas
    1.17
     starts
    1.12
    噉
    1.09
     Commencez
    1.08
    cenie
    1.07
    виси
    1.07
     estrenar
    1.07
     that
    1.05
    Activations Density 0.323%

    No Known Activations