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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 66788
    Prev
    Next
    INDEX
    Explanations

    assignments and variable declarations in programming code

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

    code syntax and structure

    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
     treinamento
    -0.48
     prestazioni
    -0.46
     unterkunft
    -0.44
     Infór
    -0.43
     excelencia
    -0.43
    Спасылкі
    -0.40
     récon
    -0.40
     Gelegenheit
    -0.40
    empereur
    -0.40
     zostało
    -0.39
    POSITIVE LOGITS
    tonsoft
    0.49
     fine
    0.44
    AlterField
    0.44
    ustand
    0.43
    parsedMessage
    0.43
    ContentAsync
    0.42
    getMinutes
    0.42
    WebServlet
    0.41
    Produ
    0.41
    viewDidLoad
    0.41
    Activations Density 0.015%

    No Known Activations