© 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
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 7457
    Prev
    Next
    INDEX
    Explanations

    reference to authors and their contributions in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    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
     kasarigan
    -0.74
     للاسماء
    -0.58
    省市镇
    -0.51
    ValueStyle
    -0.47
    TagMode
    -0.45
     विश्वसनीयता
    -0.45
     FetchType
    -0.45
     فريبيس
    -0.45
    სქოლიო
    -0.45
    ⮕
    -0.43
    POSITIVE LOGITS
     who
    0.43
     tarafından
    0.42
     whose
    0.42
    Solo
    0.42
     którzy
    0.41
     drawn
    0.40
    hObject
    0.40
    partial
    0.40
     द्वारा
    0.39
    zungs
    0.39
    Activations Density 0.437%

    No Known Activations