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. 911
    Prev
    Next
    INDEX
    Explanations

    phrases discussing the nuances of process and intent in a conceptual or analytical context

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

    elements related to intention, purpose, and reflection in writing

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

    technical form and style

    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
    rawDesc
    -0.77
     CreateTagHelper
    -0.75
    PreferredItem
    -0.75
    migrationBuilder
    -0.73
     Administrativna
    -0.69
     تضيفلها
    -0.68
    elemField
    -0.67
     ModelExpression
    -0.67
     BitmapFactory
    -0.65
    TSCA
    -0.65
    POSITIVE LOGITS
     form
    0.99
     formal
    0.89
     style
    0.89
     stylistic
    0.89
     technical
    0.85
     technique
    0.81
    form
    0.81
     Formal
    0.79
     Form
    0.78
     FORM
    0.78
    Activations Density 0.161%

    No Known Activations