Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 5620
    Prev
    Next
    INDEX
    Explanations

    concepts related to classifications and properties of sequences and materials

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

    class, compound, material, conduct, species

    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
     дописавши
    -0.28
     birlikte
    -0.27
     IV
    -0.27
    AlterField
    -0.27
    MessageTagHelper
    -0.26
     plástica
    -0.25
     dağı
    -0.25
     веке
    -0.24
     फिर
    -0.24
     dificultades
    -0.23
    POSITIVE LOGITS
     autorytatywna
    0.71
    SequentialGroup
    0.68
    RenderAtEndOf
    0.61
    
    0.60
     Administrativna
    0.60
    ſelben
    0.57
     ModelExpression
    0.55
    complexContent
    0.55
    олові
    0.54
    wezen
    0.54
    Activations Density 0.146%

    No Known Activations