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. 350
    Prev
    Next
    INDEX
    Explanations

    references to policy and policies

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

    references to official policy statements or documents.

    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
    withCredentials
    -0.45
    The
    -0.40
    Weblinks
    -0.39
     namorado
    -0.36
     dentes
    -0.36
     mystérie
    -0.36
    Find
    -0.35
     Hors
    -0.35
    from
    -0.35
    Where
    -0.35
    POSITIVE LOGITS
     POLICY
    1.34
     policy
    1.32
     Policy
    1.28
    policy
    1.24
    Policy
    1.20
    POLICY
    1.18
     policies
    1.06
     Policies
    1.05
     POLICIES
    1.01
    Policies
    0.96
    Activations Density 0.031%

    No Known Activations