© 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. 5834
    Prev
    Next
    INDEX
    Explanations

    technical terms and concepts related to research methodologies and scientific analysis

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

    scientific and research language, particularly phrases related to study methodology and academic writing.

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

    scientific paper formatting elements, such as abstracts, study descriptions, and conclusions.

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

    formal scientific-abstract discourse, especially first-person statements describing methods, results, comparisons, and study framing

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

    academic or scientific text, particularly in research papers and technical documents that describe methodologies, present findings, or discuss experimental results.

    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
     <=",
    -0.42
     numberOfRows
    -0.42
    ⌃
    -0.41
    Попис
    -0.41
    unless
    -0.40
     вот
    -0.39
     UU
    -0.39
    Hvem
    -0.39
    anything
    -0.38
    sst
    -0.38
    POSITIVE LOGITS
     Signalez
    0.65
    providedIn
    0.60
     wikipagina
    0.60
     ligiloj
    0.56
    complexContent
    0.56
     opgenomen
    0.50
     OMITTED
    0.49
    CppCodeGen
    0.48
    enumii
    0.47
    ensuremath
    0.45
    Activations Density 0.905%

    No Known Activations