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. 1202
    Prev
    Next
    INDEX
    Explanations

    positive feedback about functionality or performance

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

    works fine

    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
    ViewImports
    -0.54
     jezelf
    -0.54
    ðsíða
    -0.52
     ansatte
    -0.52
     zijne
    -0.52
     tanggung
    -0.51
    resave
    -0.50
     mijne
    -0.49
    getMenuInflater
    -0.48
    webtoken
    -0.48
    POSITIVE LOGITS
    works
    0.47
     좋
    0.46
     working
    0.46
    efficient
    0.46
    perfect
    0.45
     Worked
    0.45
     Works
    0.44
    Worked
    0.44
     Perfect
    0.43
     perfe
    0.43
    Activations Density 0.009%

    No Known Activations