© 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 - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 41782
    Prev
    Next
    INDEX
    Explanations

    reference markers or end-of-sentence indicators in structured format

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

    diverse languages and concepts

    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-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
    webElementGuid
    -0.69
     informée
    -0.67
     lenker
    -0.66
    tonsoft
    -0.65
     kasarigan
    -0.65
    SequentialGroup
    -0.64
     propOrder
    -0.58
    FieldBuilder
    -0.57
     يتيمه
    -0.57
     المعيارى
    -0.56
    POSITIVE LOGITS
    interopRequire
    0.34
     avond
    0.28
    UseVisualStyle
    0.28
     بروز
    0.28
     güne
    0.28
     beneficios
    0.27
    yorsunuz
    0.27
    皮
    0.27
     SwitchCompat
    0.27
     nghị
    0.27
    Activations Density 0.306%

    No Known Activations