© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 76177
    Prev
    Next
    INDEX
    Explanations

    municipal energy Act**Reasoning:*** **MAX_ACTIVATING_TOKENS**: Contains "Energy" and "municipal". Also contains numbers like "0", "1", "2" and "$".* **TOKENS_AFTER_MAX_ACTIVATING_TOKEN**: Contains "Act" following "0", "routes" following "municipal", and "system" following "computer".* **TOP_ACTIVATING_TEXTS**: Examples include "municipalities of Frontenac County.", "Green Energy Act", "municipal routes", "energy bills". These clearly link the concepts.* **TOP_POSITIVE_LOGITS**: Contains "Upload", "Terms", "objectives" which are general concept words, less specific to the core pattern identified.The most prominent and consistent pattern across the texts and tokens is the association between municipal entities and energy-related services or legislation (like "Energy Act"). The numerical tokens might relate to quantities or identifiers within these contexts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ברה
    -0.87
     publick
    -0.85
    phrodite
    -0.77
     秘
    -0.77
    драт
    -0.73
    Accessible
    -0.73
     zone
    -0.72
     VIVO
    -0.72
     Kais
    -0.72
     comportamento
    -0.72
    POSITIVE LOGITS
     Badger
    0.74
     Upload
    0.72
    Badger
    0.69
     upload
    0.69
    androidx
    0.66
    トレー
    0.65
    icali
    0.65
     objectives
    0.65
     Terms
    0.65
     sprawia
    0.65
    Activations Density 0.014%

    No Known Activations