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-2-2B
    4. Residual Stream - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 901
    Prev
    Next
    INDEX
    Explanations

    numerical patterns and references to numerical values in data

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

    numerical sequences or specific formats in text

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

    mathematical calculations and specific numbers

    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-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
     disambiguazione
    -0.94
    Personensuche
    -0.90
    IsMutable
    -0.80
    出版年
    -0.78
     EconPapers
    -0.78
    bootstrapcdn
    -0.77
     myſelf
    -0.76
     itſelf
    -0.76
     disponibilités
    -0.76
    XmlAccessorType
    -0.76
    POSITIVE LOGITS
     x
    0.51
    setViewportView
    0.49
    Varint
    0.48
     Green
    0.47
    x
    0.47
     Let
    0.45
     Park
    0.44
    Let
    0.44
     let
    0.43
     traditional
    0.43
    Activations Density 0.026%

    No Known Activations