© 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. 36226
    Prev
    Next
    INDEX
    Explanations

    technical or academic terms

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

    The neuron activates on tokens that are standalone numbers or alphanumeric version identifiers (e.g. “7,” “187,” “181,” “g2,” etc.).

    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
    Χ
    -1.15
     noires
    -1.15
     raiſ
    -1.13
     monté
    -1.09
     Loja
    -1.09
     vendu
    -1.05
     appétit
    -1.05
     zamiast
    -1.05
     achet
    -1.05
     classique
    -1.05
    POSITIVE LOGITS
     in
    1.26
    INTRODUCCIÓN
    1.20
     tendrán
    1.09
    外部連結
    1.00
    TRIBUN
    0.98
     すぎる
    0.96
     älteren
    0.96
    CONCLUSIONES
    0.95
    OBJETIVOS
    0.94
    section
    0.94
    Activations Density 0.123%

    No Known Activations