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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 10941
    Prev
    Next
    INDEX
    Explanations

    high-frequency patterns or entities in a specific context, particularly related to special characters or symbols

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    with diacritics, especially umlauts

    np_acts-logits-general · gemini-2.0-flash

    Swedish and German words with umlauts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ⓧ
    -0.67
    Datuak
    -0.64
     none
    -0.56
    AlterField
    -0.49
    jspx
    -0.49
    ))[
    -0.49
    dientemente
    -0.48
     AssemblyProduct
    -0.48
    tahu
    -0.46
    activity
    -0.46
    POSITIVE LOGITS
    yssey
    0.63
    óż
    0.55
     Neale
    0.54
     Majefty
    0.54
    rily
    0.53
     Babyl
    0.53
     Wraith
    0.52
     Apothe
    0.51
    riculture
    0.51
     Indah
    0.51
    Activations Density 0.097%

    No Known Activations