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. 57
    Prev
    Next
    INDEX
    Explanations

    terms related to principal component analysis (PCA) and data analysis in biological contexts

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

    Principal component analysis (PCA)

    np_acts-logits-general · gemini-2.0-flash

    principal component analysis (PCA)

    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
    ViewFeatures
    -0.71
     CiNii
    -0.61
    ativistic
    -0.57
     celib
    -0.56
     Referential
    -0.54
    Ruff
    -0.54
    GraphicsUnit
    -0.53
     GenerationType
    -0.53
     civile
    -0.51
     Registers
    -0.51
    POSITIVE LOGITS
     principal
    0.82
    Principal
    0.81
     Principal
    0.79
     cluster
    0.79
     clustering
    0.79
    principal
    0.79
    cluster
    0.75
     clusters
    0.73
     PCA
    0.71
     Cluster
    0.69
    Activations Density 0.793%

    No Known Activations