© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 5128
    Prev
    Next
    INDEX
    Explanations

    references to book excerpts and summaries

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

    instructions or rules for a structured activity (like a book meme or blog post).

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered 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
    +#+
    -0.54
    MessageTagHelper
    -0.51
    Autoritní
    -0.50
    ſelf
    -0.49
     Chriftian
    -0.49
    ſtand
    -0.45
    ſelves
    -0.44
     newName
    -0.44
     prohibido
    -0.44
    tableFuture
    -0.43
    POSITIVE LOGITS
     chapter
    0.48
     chapters
    0.45
     snip
    0.45
     snippet
    0.45
     snippets
    0.43
     sni
    0.42
     clip
    0.41
     cut
    0.41
     quote
    0.40
     grab
    0.40
    Activations Density 0.518%

    No Known Activations