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. 91289
    Prev
    Next
    INDEX
    Explanations

    references to safety features in vehicles and child equipment

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

    headrest or restraint

    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
     Paglinawan
    -0.37
    mazioni
    -0.35
    ariales
    -0.35
     Савез
    -0.35
    rolling
    -0.33
     onbe
    -0.33
    Điện
    -0.33
    diging
    -0.31
    ALLED
    -0.31
    opend
    -0.30
    POSITIVE LOGITS
     myſelf
    0.82
     tranſ
    0.76
     itſelf
    0.70
     raiſ
    0.69
     pleaſure
    0.68
    webElementGuid
    0.65
     ſte
    0.63
     fubject
    0.61
     unſ
    0.60
     viſ
    0.59
    Activations Density 0.004%

    No Known Activations