© 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. 31-GEMMASCOPE-RES-16K
    11. 14711
    Prev
    Next
    INDEX
    Explanations

    specific terms or phrases associated with scientific or technical processes

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

    numbers in scientific or technical contexts.

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

    numerical values and measurements in technical and scientific text.

    oai_token-act-pair · claude-4-5-haikuTriggered by @emiglarou

    numeric sequences or measurements in scientific and technical contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    mentions of numerical data and measurements, especially quantities, dates/years, ratios, percentages, and values paired with units (time, temperature, dimensions).

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 31-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_31/width_16k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    égalité
    -0.44
     Gefähr
    -0.43
    Tikang
    -0.43
     Fellowship
    -0.42
    НИК
    -0.41
     sizi
    -0.40
     Fingers
    -0.40
     Gewalt
    -0.40
    établissement
    -0.39
    ubles
    -0.38
    POSITIVE LOGITS
     GenerationType
    0.65
    tvguidetime
    0.64
    protoimpl
    0.54
    HtmlAttribute
    0.54
    SharedDtor
    0.54
    ReusableCell
    0.52
     sumpay
    0.50
    SBATCH
    0.49
    styleType
    0.49
     betweenstory
    0.47
    Activations Density 1.957%

    No Known Activations