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. 5778
    Prev
    Next
    INDEX
    Explanations

    programming constructs and syntax elements associated with code

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

    code constructs and assignments

    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
     GenerationType
    -0.48
    windowFixed
    -0.47
     algu
    -0.45
    DeleteBehavior
    -0.45
    IsInitialized
    -0.43
    Portale
    -0.43
     informée
    -0.43
    RegistryLite
    -0.43
     HasFactory
    -0.43
    ❉
    -0.41
    POSITIVE LOGITS
     AssemblyTitle
    0.49
    Kjelder
    0.42
    ScopeManager
    0.42
    更
    0.39
     closer
    0.36
     AssemblyProduct
    0.36
    RTSC
    0.36
    Hauptartikel
    0.35
     easier
    0.35
     ویکی‌پدی
    0.35
    Activations Density 0.032%

    No Known Activations