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. 3199
    Prev
    Next
    INDEX
    Explanations

    confirmation phrases related to deletion or removal actions

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

    Code related to deletion or removal

    np_acts-logits-general · gemini-2.0-flash

    delete confirmation

    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
    ToAdd
    -0.50
    uckoo
    -0.50
    ίων
    -0.48
     StatefulWidget
    -0.47
     aggiunto
    -0.44
    TryGetValue
    -0.43
     Discovering
    -0.42
    AddNew
    -0.42
    assembling
    -0.41
    ContentAsync
    -0.40
    POSITIVE LOGITS
     confirmation
    1.06
     Confirmation
    1.05
     delete
    1.05
     Delete
    1.04
    Confirmation
    1.01
     DELETE
    1.00
    confirmation
    0.99
    confirm
    0.99
    Delete
    0.98
     deletion
    0.98
    Activations Density 0.080%

    No Known Activations