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. 108
    Prev
    Next
    INDEX
    Explanations

    references to specific topics or entities in a structured format, such as identification phrases or definitions

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

    technical and mathematical text segments, particularly those starting with "For" or mathematical formulas.

    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
     ujednoznacz
    -0.98
    <unused68>
    -0.94
    <unused51>
    -0.93
    <unused74>
    -0.93
    <unused14>
    -0.93
    <unused8>
    -0.93
    <unused28>
    -0.93
    <unused41>
    -0.93
    <unused23>
    -0.93
    [@BOS@]
    -0.93
    POSITIVE LOGITS
    In
    0.34
     EClass
    0.31
    The
    0.29
    Self
    0.29
    "…
    0.28
    On
    0.28
    “…
    0.27
     In
    0.27
     В
    0.26
     З
    0.26
    Activations Density 0.643%

    No Known Activations