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-2-2B
    3. 21-GEMMASCOPE-TRANSCODER-16K
    4. 11047
    Prev
    Next
    INDEX
    Explanations

    punctuation and conjunctions used in dialog or related to someone telling the truth

    oai_token-act-pair · gemini-2.0-flash

    informal greetings or terms of endearment

    np_token-act-pair-logits · gpt-4o-mini

    comma

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope/layer_21/width_16k
    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
    awtextra
    -0.70
    WebServlet
    -0.66
    ​—
    -0.53
    CloseOperation
    -0.52
    DeleteBehavior
    -0.50
    FontOfSize
    -0.50
    脚注の使い方
    -0.49
     TextInputType
    -0.49
    BitmapFactory
    -0.48
    makeText
    -0.48
    POSITIVE LOGITS
     dude
    1.33
     mate
    1.27
     sir
    1.20
     buddy
    1.20
     dear
    1.17
     guys
    1.13
     folks
    1.13
     darling
    1.09
     gentlemen
    1.09
     brother
    1.08
    Activations Density 1.417%

    No Known Activations