© 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. 1215
    Prev
    Next
    INDEX
    Explanations

    occurrences of specific historical events and timelines

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

    dates and years, especially those in the late 19th and early 20th centuries.

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

    phrases marking the end of a period or a historical transition, typically tied to a specific year and describing cessation, replacement, or retirement.

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

    historical time periods ending or transitioning, particularly indicated by "until" followed by a year and phrases describing what happened when that period ended.

    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
    Autoritní
    -0.52
     ModelExpression
    -0.45
     pica
    -0.41
     benefit
    -0.38
     diana
    -0.38
     perspective
    -0.37
    arusian
    -0.37
     topic
    -0.37
     Cicero
    -0.36
     caffe
    -0.36
    POSITIVE LOGITS
    principalColumn
    0.63
     विश्वसनीयता
    0.52
     letzt
    0.50
    migrationBuilder
    0.46
     losing
    0.46
    centralwidget
    0.45
     دیکھیے
    0.45
    letzt
    0.44
    PRNewswire
    0.44
     superseded
    0.44
    Activations Density 0.385%

    No Known Activations