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 - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 366
    Prev
    Next
    INDEX
    Explanations

    references to decision-making processes and guidelines

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

    phrases and words related to providing advice or guidance.

    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
     tartalomajánló
    -0.73
    GEBURTSDATUM
    -0.72
     Савезне
    -0.72
    Билгалдахарш
    -0.71
     Chwiliwch
    -0.71
     مشين
    -0.71
     Paglinawan
    -0.70
     Numerade
    -0.69
     Taktlose
    -0.69
     ComVisible
    -0.68
    POSITIVE LOGITS
    ↵↵
    0.40
    <strong>
    0.37
    <h1>
    0.37
    er
    0.36
    b
    0.35
     details
    0.35
    ary
    0.35
     based
    0.35
     tactics
    0.34
    i
    0.34
    Activations Density 0.338%

    No Known Activations