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. 24068
    Prev
    Next
    INDEX
    Explanations

    elements related to HTTP request headers and their structure

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

    API request formatting

    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-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
    textview
    -0.38
    delwed
    -0.37
     cu
    -0.35
    morrow
    -0.33
    aspectj
    -0.33
     agrí
    -0.33
    SqlServer
    -0.32
     getModel
    -0.32
    ViewPager
    -0.31
     μον
    -0.31
    POSITIVE LOGITS
     المعيارى
    0.72
     informée
    0.68
     BoxDecoration
    0.66
     InputDecoration
    0.65
     invokingState
    0.63
    tagHelperRunner
    0.62
    ArrowToggle
    0.62
     sumpay
    0.56
    ViewFeatures
    0.55
    getHeaders
    0.55
    Activations Density 0.017%

    No Known Activations