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-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 15266
    Prev
    Next
    INDEX
    Explanations

    legal terminology and formal agreements or statements

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

    Statements, letters, or memos

    np_acts-logits-general · gemini-2.0-flash

    stating or documenting communications

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
     BorderRadius
    -0.60
    "]);
    
    -0.60
    }")
    
    -0.58
    endphp
    -0.58
    Бахар
    -0.57
     AssemblyTitle
    -0.57
    culable
    -0.56
    Viitteet
    -0.56
    ]})
    -0.55
    centralwidget
    -0.54
    POSITIVE LOGITS
     stating
    1.56
     indicating
    1.54
     saying
    1.42
     explaining
    1.40
     showing
    1.39
     describing
    1.37
     outlining
    1.34
     telling
    1.34
     detailing
    1.32
     suggesting
    1.31
    Activations Density 0.902%

    No Known Activations