© 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. 11050
    Prev
    Next
    INDEX
    Explanations

    formal legal and technical references

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

    formal, structured text elements—requirements/definitions, legal or technical headings, and their surrounding punctuation (quotes, parentheses, colons, periods) indicating specification-style statements.

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

    Looking at the activations, this neuron activates strongly on: 1. Quotation marks (especially closing quotes with periods: `."` and `)` showing very high activations) 2. Formal/technical language patterns, particularly in legal, technical specifications, and documentation

    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
     виправивши
    -0.47
    OGND
    -0.44
     estekak
    -0.42
    تقاوى
    -0.40
    ########.
    -0.37
    #+#
    -0.37
    äsentation
    -0.36
     دیکھیے
    -0.36
    endphp
    -0.36
     bağlantılar
    -0.36
    POSITIVE LOGITS
    fjspx
    0.62
     betweenstory
    0.60
    RegistryLite
    0.59
     initComponents
    0.51
     juist
    0.50
    ceğini
    0.49
     Coolidge
    0.47
     Bluff
    0.46
    bbero
    0.46
    googleapis
    0.46
    Activations Density 0.476%

    No Known Activations