Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 225
    Prev
    Next
    INDEX
    Explanations

    cleaning actions and substances

    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
    dibles
    -1.02
     sufren
    -1.00
    webElement
    -0.97
     explore
    -0.93
    楽しみです
    -0.92
     Anyone
    -0.91
    앓
    -0.91
     reciben
    -0.89
     Glied
    -0.89
     Whoever
    -0.88
    POSITIVE LOGITS
    cleaning
    1.58
     Cleaning
    1.47
    Cleaning
    1.45
     cleaning
    1.41
     CLEANING
    1.40
     cleans
    1.30
    🧼
    1.29
     limpieza
    1.28
     cleaned
    1.28
    CLEAN
    1.20
    Activations Density 0.016%

    No Known Activations