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. 15671
    Prev
    Next
    INDEX
    Explanations

    technical terms and mathematical concepts related to statistical models and estimations

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

    Comes before math symbols or citations

    np_acts-logits-general · gemini-2.0-flash

    dynamic discrete choice

    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
    rawtypes
    -0.45
    haup
    -0.42
    unnitel
    -0.42
    nsan
    -0.42
    وفاته
    -0.40
    opc
    -0.40
     Beres
    -0.40
    wyr
    -0.40
     Palmer
    -0.39
     🤷
    -0.39
    POSITIVE LOGITS
     firms
    0.91
     Firms
    0.84
     firm
    0.83
     exogenous
    0.81
    CodedInputStream
    0.80
     endogenous
    0.78
     shocks
    0.77
     EconPapers
    0.72
     defaultstate
    0.71
     Gennaio
    0.71
    Activations Density 2.361%

    No Known Activations