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-2-2B
    4. Residual Stream - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 869
    Prev
    Next
    INDEX
    Explanations

    keywords related to specific types of food and culinary items

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

    references to various types of food, particularly hamburgers and pancakes

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

    food items and their makers

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
    Erfolge
    -0.59
    awtextra
    -0.51
     persoane
    -0.47
    quested
    -0.46
     saker
    -0.46
     jScrollPane
    -0.45
     Meksiku
    -0.45
    εια
    -0.45
    Jeografia
    -0.44
     parfüm
    -0.44
    POSITIVE LOGITS
     maker
    0.81
     <<<<<<<<<<<<<<
    0.72
     makers
    0.71
    medriver
    0.65
     povezave
    0.64
     IFTTT
    0.64
     Maker
    0.62
    ADELPHIA
    0.61
     AppComponent
    0.61
    ollectionView
    0.60
    Activations Density 0.054%

    No Known Activations