© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 14397
    Prev
    Next
    INDEX
    Explanations

    instances of code or formatting in a markup language

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

    LaTeX preamble and document structure

    np_acts-logits-general · gemini-2.0-flash

    LaTeX and HTML code

    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
    ngrx
    -0.82
    :✨
    -0.81
    RenderAtEndOf
    -0.76
     bezeichneter
    -0.75
     queſta
    -0.71
    AndEndTag
    -0.69
    ſſung
    -0.69
     unſer
    -0.68
     AssemblyCulture
    -0.68
     InputDecoration
    -0.67
    POSITIVE LOGITS
     الحره
    0.38
     digitais
    0.36
     olvides
    0.34
     difficiles
    0.31
     and
    0.31
     dependencia
    0.30
     communs
    0.29
     especiais
    0.29
     principalement
    0.29
     particulières
    0.28
    Activations Density 0.011%

    No Known Activations