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. 10918
    Prev
    Next
    INDEX
    Explanations

    terms associated with art, specifically related to artists and their works

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

    ending in "uk", "um", "utt", "uc", or "ux"

    np_acts-logits-general · gemini-2.0-flash

    German words, compound 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
    ValueStyle
    -0.74
    GenerationType
    -0.71
     /*#__
    -0.69
    DotNetBar
    -0.63
    windowFixed
    -0.61
     freshest
    -0.60
     GenerationType
    -0.60
    noons
    -0.59
     truest
    -0.59
     Wicidata
    -0.58
    POSITIVE LOGITS
     Nuss
    0.71
     Audiodateien
    0.66
    ünkü
    0.59
    velopes
    0.59
    fördert
    0.59
    probieren
    0.58
     rekomendasi
    0.58
    chtigkeit
    0.57
    uuuu
    0.57
    ubus
    0.56
    Activations Density 1.035%

    No Known Activations