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. 3-GEMMASCOPE-RES-16K
    6. 11164
    Prev
    Next
    INDEX
    Explanations

    formatted text elements such as bold, italic, and code snippets in the document

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.3.hook_resid_post
    Hook Layer
    3
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ANTLR
    -0.72
    Abraço
    -0.60
    LikeLiked
    -0.58
    équ
    -0.56
     utafiti
    -0.56
    zugs
    -0.55
    wirk
    -0.55
     haft
    -0.54
    гла
    -0.54
     dét
    -0.54
    POSITIVE LOGITS
    <i>
    1.06
    <em>
    1.02
    <u>
    1.00
    <strong>
    0.96
    <b>
    0.95
    MessageTagHelper
    0.83
     Chwiliwch
    0.82
    __(/*!
    0.80
    <s>
    0.76
    <sub>
    0.74
    Activations Density 0.003%

    No Known Activations