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. 2411
    Prev
    Next
    INDEX
    Explanations

    specific biochemical compounds and their interactions in cellular processes

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

    or tokens before capitalized words

    np_acts-logits-general · gemini-2.0-flash

    combining parts of names and terms

    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
     跳转至
    -0.62
    elsif
    -0.56
     mää
    -0.55
     vastaan
    -0.54
    <tbody>
    -0.53
     tää
    -0.53
     enää
    -0.52
    mær
    -0.50
    __;
    -0.49
     mahdol
    -0.49
    POSITIVE LOGITS
     يتيمه
    1.05
    principalColumn
    0.87
    IsMutable
    0.83
     kaynağından
    0.82
    下载附件
    0.81
    IndentedString
    0.76
     ComVisible
    0.74
    fjspx
    0.73
    adpleegd
    0.72
    GeoNames
    0.72
    Activations Density 1.426%

    No Known Activations