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. 10329
    Prev
    Next
    INDEX
    Explanations

    sentences or segments that reference significant events, actions, or statements related to political or military issues

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

    Comes before the name or title of a person

    np_acts-logits-general · gemini-2.0-flash

    critics and lawmakers

    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
    acyj
    -0.61
     nữa
    -0.60
    InstrumentedTest
    -0.59
    ViewImports
    -0.57
     itemName
    -0.56
     malheureusement
    -0.55
    +#+#
    -0.55
    JdbcTemplate
    -0.54
    ViewFeatures
    -0.54
    IsContent
    -0.53
    POSITIVE LOGITS
    Critics
    0.98
     critics
    0.92
     Earlier
    0.91
     Critics
    0.89
     analysts
    0.88
     President
    0.87
    Analysts
    0.87
     advocates
    0.85
    Earlier
    0.83
     lawmakers
    0.82
    Activations Density 0.352%

    No Known Activations