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-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 92129
    Prev
    Next
    INDEX
    Explanations

    references to the name "Sandy" and the concept of luck

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

    names with -y suffix

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     vandens
    -0.59
     BoxDecoration
    -0.55
     vodu
    -0.54
     dům
    -0.54
     tričko
    -0.53
    Билгалдахарш
    -0.51
     strøm
    -0.51
     cellulare
    -0.51
    rungsseite
    -0.50
    fahan
    -0.50
    POSITIVE LOGITS
     Rocky
    0.73
    rocky
    0.71
     Sandy
    0.69
    rusty
    0.66
    handy
    0.66
     Moody
    0.64
    rainy
    0.64
    eky
    0.63
    Leaky
    0.63
     Snowy
    0.62
    Activations Density 0.006%

    No Known Activations