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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 5286
    Prev
    Next
    INDEX
    Explanations

    the start of a document or significant sections within it

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

    Code snippets, symbols, and uncommon words

    np_acts-logits-general · gemini-2.0-flash

    technical jargon and foreign words

    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
     kasarigan
    -0.56
    thansa
    -0.52
     समीक्षक
    -0.51
     propOrder
    -0.49
    freunde
    -0.48
    riwal
    -0.47
     betweenstory
    -0.47
     étrangères
    -0.47
    خصة
    -0.47
     Griechen
    -0.45
    POSITIVE LOGITS
    tagHelperRunner
    1.26
     (!__
    0.78
     وتسجيلات
    0.77
     ujednoznacz
    0.74
    XtraGrid
    0.72
    UnusedPrivate
    0.66
    ftagPool
    0.64
    Autoritní
    0.62
    tvguidetime
    0.62
     otomatig
    0.62
    Activations Density 0.042%

    No Known Activations