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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 1885
    Prev
    Next
    INDEX
    Explanations

    numerical data and statistical results in a context related to financial or medical assessments

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

    numerical data and statistical indicators, like probabilities, percentages, and scientific notation.

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

    entries of numerical measurements in scientific/tabular formats, including decimals, symbols, ranges, and missing-value notation common to LaTeX-style data.

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

    numerical data in tabular or statistical formats, particularly decimal values in tables with aligned columns.

    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
    delwed
    -0.33
    memoized
    -0.32
    pushFollow
    -0.31
     kenyataan
    -0.31
     ihtiyac
    -0.31
    MMdd
    -0.30
     Vikipedi
    -0.28
     memakan
    -0.28
    getOriginal
    -0.28
    gelöst
    -0.27
    POSITIVE LOGITS
     nahilalakip
    0.70
     transfieras
    0.59
     zwiſchen
    0.57
    ябре
    0.56
    findpost
    0.55
    ⸄
    0.53
     
    0.52
    Formik
    0.51
    tonsoft
    0.50
    ◚
    0.50
    Activations Density 0.393%

    No Known Activations