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. 4698
    Prev
    Next
    INDEX
    Explanations

    references to specific context words indicating reiteration or clarification

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

    "this" followed by certain nouns

    np_acts-logits-general · gemini-2.0-flash

    this way or context

    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
     חיצוניים
    -0.65
    InjectAttribute
    -0.57
    出版年
    -0.56
     disambiguazione
    -0.56
     ویکی‌پدی
    -0.54
    الإنجليزية
    -0.53
    writeFieldEnd
    -0.53
     accanto
    -0.52
     ajuns
    -0.52
     seguintes
    -0.51
    POSITIVE LOGITS
     regard
    1.19
    regard
    0.89
     respect
    0.84
     vein
    0.82
     regards
    0.80
     REGARD
    0.72
     respecto
    0.70
     sense
    0.70
     context
    0.69
     vain
    0.66
    Activations Density 0.161%

    No Known Activations