© 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. 2552
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions involving inequalities and conditions

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

    mathematical or programming syntax like symbols, variables, and code-related tokens.

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

    structured technical/markup tokens—especially LaTeX math delimiters and expressions, bracketed citations, and code/JSON syntax markers.

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

    line breaks or newlines in structured code and data formats.

    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
    .
    -0.47
    ".
    -0.41
    '.
    -0.41
    !
    -0.40
    ').
    -0.40
    ».
    -0.40
    ").
    -0.39
    <eos>
    -0.39
     :
    -0.38
    "".
    -0.38
    POSITIVE LOGITS
     חיצוניים
    0.94
     nahilalakip
    0.89
     autorytatywna
    0.83
    jsxFileName
    0.74
    tvguidetime
    0.71
    awtextra
    0.69
     שוליים
    0.69
    StructEnd
    0.68
     Taktlose
    0.66
    verwijspagina
    0.65
    Activations Density 0.574%

    No Known Activations