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. 6641
    Prev
    Next
    INDEX
    Explanations

    elements related to crafting and decorative embellishments

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

    Added embellishments or decorations

    np_acts-logits-general · gemini-2.0-flash

    added for decoration

    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
     ujednoznacz
    -0.97
     CreateTagHelper
    -0.96
    DrawerToggle
    -0.70
    ViewImports
    -0.69
    IntoConstraints
    -0.69
    setVerticalGroup
    -0.67
    ⓧ
    -0.67
     quæ
    -0.65
     disambiguazione
    -0.64
     calldata
    -0.64
    POSITIVE LOGITS
     embell
    0.71
     decorating
    0.60
     decoration
    0.60
     decorations
    0.59
     decorate
    0.58
    rofoam
    0.58
     decorative
    0.57
     שוליים
    0.54
     decou
    0.53
     glitter
    0.53
    Activations Density 0.135%

    No Known Activations