© 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. 1081
    Prev
    Next
    INDEX
    Explanations

    scientific terminology and formatting consistent with academic research papers

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

    statistical and scientific notation terms like p-values, exponents, and measurement units.

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

    scientific notation and units of measurement in academic papers.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi
    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
     betweenstory
    -0.82
     nahilalakip
    -0.67
     незавершена
    -0.62
    outSlope
    -0.62
    uxxxx
    -0.60
    SharedDtor
    -0.59
     مرئيه
    -0.57
     defaultstate
    -0.57
     виправивши
    -0.57
    ragalactic
    -0.57
    POSITIVE LOGITS
     DbType
    0.40
    alei
    0.40
    øde
    0.39
    IVA
    0.38
    Locale
    0.37
    ITY
    0.37
     Cameron
    0.37
     works
    0.37
    LookAnd
    0.36
    PRNewswire
    0.36
    Activations Density 0.934%

    No Known Activations