© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 6139
    Prev
    Next
    INDEX
    Explanations

    corporate and scientific references

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

    technical scientific document metadata, particularly company names and scientific instrument details.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi

    references to companies and institutions (including trademarks, corporate suffixes, locations, dates, and catalog/model identifiers) typical of scientific methods and commercial product descriptions.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    company names, brand names, and corporate identifiers in technical and commercial text.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    Prompts (Dashboard)
    24,576 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
     dAtA
    -0.63
     autorytatywna
    -0.62
    astéroïdes
    -0.56
    TagMode
    -0.56
    ----</
    -0.55
    oredCriteria
    -0.54
     <=",
    -0.54
     Infórmanos
    -0.53
     تضيفلها
    -0.51
    Datuak
    -0.49
    POSITIVE LOGITS
    setBold
    0.48
    engagent
    0.35
    ContentAsync
    0.34
     nahilalakip
    0.32
     processing
    0.32
    Slf
    0.31
    исленность
    0.31
     Processing
    0.31
    المكان
    0.31
    انگلیسی
    0.31
    Activations Density 0.144%

    No Known Activations