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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 21-GEMMASCOPE-RES-16K
    6. 5832
    Prev
    Next
    INDEX
    Explanations

    references to examples or illustrative cases in the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.hook_resid_post
    Hook Layer
    21
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Godhead
    -0.42
    thschild
    -0.42
    dR
    -0.41
    𝖉
    -0.41
    <strong>
    -0.40
    retention
    -0.40
    noir
    -0.40
     McNally
    -0.40
     utara
    -0.40
            
    -0.39
    POSITIVE LOGITS
     Example
    1.20
    Example
    1.20
    example
    1.18
     example
    1.17
     examples
    1.13
    EXAMPLE
    1.10
     EXAMPLE
    1.09
    Examples
    1.00
     Examples
    0.99
    examples
    0.98
    Activations Density 0.183%

    No Known Activations