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. 991
    Prev
    Next
    INDEX
    Explanations

    technical terms related to programming and library management

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

    programming and configuration-related terms in a technical context

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

    linker flags or options

    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
     CURIAM
    -0.80
    archiviato
    -0.71
     Профиль
    -0.67
    niająca
    -0.66
     AssemblyProduct
    -0.65
    IsContent
    -0.64
     ModelRenderer
    -0.64
     forceps
    -0.64
    èdia
    -0.63
    LabelTagHelper
    -0.63
    POSITIVE LOGITS
     -
    0.58
    alez
    0.47
     –
    0.46
    gnato
    0.44
    framework
    0.43
     resa
    0.42
    marle
    0.39
    Dor
    0.39
    фр
    0.39
     prisa
    0.37
    Activations Density 0.115%

    No Known Activations