© 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. 1382
    Prev
    Next
    INDEX
    Explanations

    coding structures and data attributes related to programming

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

    programming syntax and code fragments across multiple programming languages and frameworks (HTML, Android XML, TypeScript, mathematical notation, etc.).

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

    end-of-construct syntax in code and markup, i.e., tokens that close or finalize structured elements (tags, braces, bracketed/parenthesized statements).

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

    elements or variables being accessed by index within arrays or data structures in programming code.

    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
     незавершена
    -0.49
    yntaxException
    -0.46
    tagHelperRunner
    -0.41
    XmlEnum
    -0.41
     виправивши
    -0.41
    balleur
    -0.41
     препратки
    -0.41
    setVerticalGroup
    -0.40
    Havolalar
    -0.40
     Réponses
    -0.39
    POSITIVE LOGITS
    AutoScaleMode
    0.54
    galan
    0.46
    piller
    0.46
    sequential
    0.44
     concor
    0.43
    cipated
    0.43
    orical
    0.43
    chard
    0.43
     Browser
    0.42
     Psi
    0.42
    Activations Density 0.644%

    No Known Activations