© 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-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 211
    Prev
    Next
    INDEX
    Explanations

    structured data related to coding or programming contexts

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

    structured data fields, labels, and values typically found in code snippets, tables, or serialized data formats such as JSON or database outputs.

    oai_token-act-pair · gpt-4.1-2025-04-14Triggered by @ywhp2c8tjm

    code snippets and URLs as well as references to academic papers and also some random terms that appear in technical writing.

    oai_token-act-pair · gemini-2.0-flashTriggered by @ywhp2c8tjm

    Tokens preceding bracketed citations

    np_acts-logits-general · gemini-2.0-flash

    leading to, accounting for

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    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
    tagHelperRunner
    -0.97
    AddTagHelper
    -0.76
     Administrativna
    -0.74
     CreateTagHelper
    -0.71
    TagMode
    -0.66
    RenderAtEndOf
    -0.63
    AndEndTag
    -0.56
    ])));
    -0.56
     Мексичка
    -0.54
     EnglishChoose
    -0.53
    POSITIVE LOGITS
    0.36
    attn
    0.31
    ­
    0.29
    brew
    0.28
    ATT
    0.28
    ла
    0.28
    uchy
    0.28
    Fed
    0.28
    aka
    0.27
    istico
    0.27
    Activations Density 86.057%

    No Known Activations