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 - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 829
    Prev
    Next
    INDEX
    Explanations

    references to a specific cultural or geographical term, particularly related to "Jal," which may denote a place or concept

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

    mentions of a specific village

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

    jalapeno

    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-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
     ویکی‌پدیای
    -0.59
     Festa
    -0.51
    aventura
    -0.50
     câte
    -0.49
     blaster
    -0.46
     practice
    -0.45
    Practice
    -0.45
    وذ
    -0.45
     fechar
    -0.45
    Nau
    -0.44
    POSITIVE LOGITS
    Datuak
    0.91
    Obrázky
    0.76
     ProtoMessage
    0.66
    tagHelperRunner
    0.65
     kasarigan
    0.63
    ("]");
    0.61
    Pyx
    0.60
    devamını
    0.58
     arşivlendi
    0.57
    RectangleBorder
    0.56
    Activations Density 0.046%

    No Known Activations