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 - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 7735
    Prev
    Next
    INDEX
    Explanations

    phrases and terms related to storage, preservation, and effective packing techniques

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

    "Storage" or "temporary" followed by a related concept

    np_acts-logits-general · gemini-2.0-flash

    storage

    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-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    новниш
    -0.55
     Penetration
    -0.50
     Oedipus
    -0.49
    dymyr
    -0.48
     penetration
    -0.47
    brecht
    -0.45
     Perci
    -0.45
    vaadin
    -0.45
     vru
    -0.44
    protoc
    -0.44
    POSITIVE LOGITS
     storage
    2.75
     Storage
    2.40
     STORAGE
    2.26
     storing
    2.25
    storage
    2.23
    Storage
    2.19
     stored
    2.15
     Storing
    1.97
    STORAGE
    1.96
    Storing
    1.90
    Activations Density 0.366%

    No Known Activations