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. 14354
    Prev
    Next
    INDEX
    Explanations

    references to the concept of degeneracy in various contexts

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

    degenerate or nuanced

    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
    show
    -0.47
    ('*
    -0.46
    star
    -0.46
    ujudkan
    -0.45
    test
    -0.45
    bit
    -0.42
    IS
    -0.42
    *
    -0.42
    *}
    -0.42
     Millar
    -0.41
    POSITIVE LOGITS
     autorytatywna
    0.78
    umerate
    0.71
     enumerate
    0.65
     <<<<<<<<<<<<<<
    0.60
     TextAppearance
    0.59
     PPR
    0.58
    Enumeration
    0.57
    extField
    0.56
    urlencoded
    0.56
     enumeration
    0.55
    Activations Density 0.006%

    No Known Activations