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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-2B
    7. Residual Stream - 65k
    9. 20-GEMMASCOPE-RES-65K
    11. 2687
    Prev
    Next
    INDEX
    Explanations

    the Arabic character "ك" and variations of the word "бесконечный" in Russian, indicating a focus on certain script or vocabulary

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

    initials and abbreviations

    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
     Jefus
    -0.82
     Chriſt
    -0.73
     itſelf
    -0.71
    참고
    -0.68
     ſy
    -0.63
    ſelves
    -0.61
     themſelves
    -0.61
     houſe
    -0.61
     ſever
    -0.60
     himſelf
    -0.59
    POSITIVE LOGITS
     CreateTagHelper
    0.76
    +#+#
    0.71
    UrlResolution
    0.69
     utafitiHapana
    0.67
     consultato
    0.65
    ordinaire
    0.58
     UIAlert
    0.56
     numberOfRows
    0.56
    tagHelperRunner
    0.55
     насељу
    0.54
    Activations Density 0.001%

    No Known Activations