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. 11097
    Prev
    Next
    INDEX
    Explanations

    expressions of positive feelings and experiences related to enjoyment and relaxation

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

    "wake", "woke", or "waking"

    np_acts-logits-general · gemini-2.0-flash

    getting up early

    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
     snippetHide
    -0.53
    uku
    -0.53
    HTTPClient
    -0.46
     miesię
    -0.46
     uzav
    -0.46
     samtidig
    -0.45
    terreno
    -0.44
     måned
    -0.44
    ultura
    -0.44
    oteca
    -0.44
    POSITIVE LOGITS
     waking
    1.95
     wake
    1.88
     woken
    1.83
     woke
    1.82
     wakes
    1.78
     morning
    1.76
     breakfast
    1.71
     Wake
    1.65
     awoke
    1.59
    Wake
    1.58
    Activations Density 0.197%

    No Known Activations