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. 120
    Prev
    Next
    INDEX
    Explanations

    phrases indicating the latest or first occurrences of events, developments, or features

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

    words indicating first occurrences or milestones in history.

    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
    fromnode
    -0.63
    ControllerAdvice
    -0.45
     enfans
    -0.44
     itinéraires
    -0.43
     náv
    -0.41
    bildēt
    -0.40
    ulemon
    -0.40
     olisi
    -0.39
     dlouho
    -0.39
     dessutom
    -0.38
    POSITIVE LOGITS
    enumi
    0.90
    المناصب
    0.64
     विश्वसनीयता
    0.62
    __*/
    0.60
     surla
    0.60
    Билгалдахарш
    0.57
    maining
    0.56
     two
    0.55
    enumii
    0.55
     defaultstate
    0.55
    Activations Density 0.243%

    No Known Activations