Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 73552
    Prev
    Next
    INDEX
    Explanations

    references to specific brands of footwear and related descriptors

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

    Page numbers and document identifiers in legal texts, typically appearing at the bottom or corners of pages, often preceded by dashes or followed by special formatting.

    eleuther_acts_top20 · claude-3-7-sonnet-20250219Triggered by @lu-christina
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
    地
    -0.32
    ~~~~~~~~~~~~~~~~
    -0.31
     дальней
    -0.30
    周
    -0.30
     sok
    -0.30
    台
    -0.28
    ~~~~~~~~
    -0.28
    <eos>
    -0.28
    Péri
    -0.28
    Vpc
    -0.27
    POSITIVE LOGITS
     CreateTagHelper
    1.14
    rungsseite
    0.90
     autorytatywna
    0.80
    SourceChecksum
    0.77
     ujednoznacz
    0.75
    tvguidetime
    0.72
     <<<<<<<<<<<<<<
    0.72
     pinulongan
    0.72
    TagMode
    0.71
    səhifə
    0.69
    Activations Density 0.280%

    No Known Activations