© 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-2-2B
    4. Residual Stream - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 1787
    Prev
    Next
    INDEX
    Explanations

    references to historical figures and places connected to Scottish settlement in Ireland

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

    Scottish places and entities

    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
    󠁥
    -0.95
    మీ
    -0.55
    AccessorTable
    -0.50
    Iya
    -0.49
    жели
    -0.48
     tf
    -0.47
    siella
    -0.47
    |}{$
    -0.47
     intérêts
    -0.46
     chande
    -0.46
    POSITIVE LOGITS
     Scottish
    1.36
     Glasgow
    1.32
     Scotland
    1.28
    Scottish
    1.25
     Scots
    1.24
    Glasgow
    1.24
     Edinburgh
    1.21
    Scotland
    1.14
    Edinburgh
    1.07
     scottish
    1.04
    Activations Density 0.182%

    No Known Activations