© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 4805
    Prev
    Next
    INDEX
    Explanations

    instances of scientific analysis related to data interpretation and methodological rigor

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

    scientific and research-related technical terms, especially words related to data analysis, research methods, and scientific processes.

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

    scientific text related to data analysis and interpretation in academic research.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    descriptions of scientific data analysis and evaluation tasks—often in methods or author-contribution sections—covering processing, interpretation, and reporting of results, sometimes with software or lab procedures.

    oai_token-act-pair · gpt-5Triggered 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
    BagConstraints
    -0.50
    transQ
    -0.47
    posedge
    -0.46
    abestanden
    -0.45
    лтамалар
    -0.45
    wapV
    -0.44
     PRACTICE
    -0.43
    practice
    -0.41
     HERO
    -0.40
    ernoo
    -0.40
    POSITIVE LOGITS
     results
    0.55
    interprétation
    0.52
     interpretations
    0.52
     interpretation
    0.50
    results
    0.50
     hasilnya
    0.49
     resultaten
    0.48
    interpreted
    0.47
     resultados
    0.47
    Results
    0.46
    Activations Density 0.802%

    No Known Activations