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. 962
    Prev
    Next
    INDEX
    Explanations

    structured data and key-value pairs in code or configuration settings

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

    programming syntax and structural elements in code snippets

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

    key-value pairs

    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
    ENCY
    -0.57
     adultes
    -0.56
    îné
    -0.55
     "").
    -0.52
    phir
    -0.51
    اضي
    -0.50
    ถม
    -0.49
     Michaelis
    -0.49
    슷
    -0.49
     Turnpike
    -0.48
    POSITIVE LOGITS
    KommentareTeilen
    0.78
    aarrggbb
    0.67
     CreateTagHelper
    0.65
    endphp
    0.63
    grà
    0.63
    ThemeOverlay
    0.61
    таратура
    0.60
     NSCoder
    0.60
    OGND
    0.59
    évaluateur
    0.57
    Activations Density 0.052%

    No Known Activations