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. 782
    Prev
    Next
    INDEX
    Explanations

    references to academic articles or citations related to scientific research

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

    references to academic papers or citations related to arXiv

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

    arXiv identifiers

    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
    LogFactory
    -0.73
    contentLoaded
    -0.65
     sanitaires
    -0.60
    CodedInputStream
    -0.59
    énieurs
    -0.59
    ★★★★☆
    -0.58
    Étymologie
    -0.57
     réguli
    -0.57
     énergé
    -0.56
    SerializedSize
    -0.55
    POSITIVE LOGITS
    arXiv
    1.00
    
    0.67
     arXiv
    0.67
     arxiv
    0.61
     للاسماء
    0.60
     EconPapers
    0.59
    WEBPACK
    0.58
    بوابة
    0.53
    UVWXYZ
    0.53
     الاطلاع
    0.52
    Activations Density 0.005%

    No Known Activations