© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 3866
    Prev
    Next
    INDEX
    Explanations

    punctuation and certain function words that indicate emotional or opinionated expressions

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

    Sentences followed by "<end_of_turn>" token

    np_acts-logits-general · gemini-2.0-flash

    Ultimately

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 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
     Италијани
    -0.74
    bootstrapcdn
    -0.73
    Derbyniad
    -0.70
    migrationBuilder
    -0.69
     pinulongan
    -0.69
    Демографія
    -0.68
    الدراسه
    -0.67
    DockStyle
    -0.65
    IUrlHelper
    -0.64
     surla
    -0.64
    POSITIVE LOGITS
     ultimately
    0.35
    Exactly
    0.35
     thăng
    0.35
     Simply
    0.34
    Simply
    0.34
     TITLE
    0.34
    +:+
    0.33
    Sure
    0.32
    Awesome
    0.32
     Absolutely
    0.32
    Activations Density 0.162%

    No Known Activations