Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 28189
    Prev
    Next
    INDEX
    Explanations

    specific syntactical structures in programming code and mathematical expressions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
     שוליים
    -0.54
    +#+#
    -0.52
    曖昧さ回避
    -0.52
    onViewCreated
    -0.51
    
    -0.50
     resourceCulture
    -0.48
    ChildScrollView
    -0.48
    AutoScale
    -0.47
     InputDecoration
    -0.47
    richTextPanel
    -0.46
    POSITIVE LOGITS
    цездатний
    0.42
     Савезне
    0.41
    ⎨
    0.38
    hemian
    0.38
     bes
    0.35
    yntaxException
    0.35
    Checked
    0.33
    &__
    0.32
    sonder
    0.32
    يكب
    0.32
    Activations Density 0.014%

    No Known Activations