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. 4667
    Prev
    Next
    INDEX
    Explanations

    specific names or titles, particularly related to individuals and works in various contexts

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

    starting with "Fri", "Bri", "Gri", "Cri"

    np_acts-logits-general · gemini-2.0-flash

    cri, fri, gri, dri, bri

    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
    tagext
    -0.63
     ویکی‌پدیای
    -0.59
     SqlDbType
    -0.58
    Datuak
    -0.54
    хьтан
    -0.52
    اریخ
    -0.51
    atrième
    -0.51
    tation
    -0.50
     mouseY
    -0.50
     سكانية
    -0.50
    POSITIVE LOGITS
     estimés
    0.62
    Cri
    0.62
     Cri
    0.59
    cri
    0.58
    mawia
    0.55
    sidemargin
    0.55
    msgid
    0.55
     muligt
    0.53
    
    0.53
    OOTDTY
    0.52
    Activations Density 0.167%

    No Known Activations