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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 5226
    Prev
    Next
    INDEX
    Explanations

    terms related to legal and bureaucratic concepts

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

    Text after questions or statements

    np_acts-logits-general · gemini-2.0-flash

    countries and organizations

    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
    ValueStyle
    -0.67
    cientes
    -0.66
    URLException
    -0.62
    TagHelper
    -0.60
     propOrder
    -0.60
    ++]=
    -0.59
     ExecuteAsync
    -0.57
    ConstraintMaker
    -0.57
     surla
    -0.56
     LoggerFactory
    -0.55
    POSITIVE LOGITS
     USA
    0.89
     UK
    0.84
    USA
    0.78
     America
    0.77
     India
    0.71
     Canada
    0.70
     Deutschland
    0.68
     Greater
    0.66
     Australia
    0.66
    America
    0.66
    Activations Density 0.247%

    No Known Activations