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. 10769
    Prev
    Next
    INDEX
    Explanations

    geographical references and names related to specific regions

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

    regions ending in "ider" or "Baltic"

    np_acts-logits-general · gemini-2.0-flash

    river names and regions

    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
     miej
    -0.52
    spender
    -0.52
    tagHelperRunner
    -0.51
     vectorielles
    -0.50
     embolism
    -0.48
     onlyOwner
    -0.47
     visuales
    -0.47
     malheureux
    -0.46
     mengisi
    -0.46
     glyphicon
    -0.46
    POSITIVE LOGITS
     region
    1.14
     Region
    1.05
    region
    0.90
     County
    0.82
     province
    0.81
     Valley
    0.80
     Peninsula
    0.78
     Province
    0.77
     county
    0.76
     région
    0.76
    Activations Density 0.365%

    No Known Activations