© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    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. 1389
    Prev
    Next
    INDEX
    Explanations

    instances of the beginning of a new section or paragraph in text

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

    the start of new segments or sections in the text

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

    computer systems processing technical specifications

    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
     
    -0.66
    ↵
    -0.66
    .
    -0.66
     (
    -0.63
     F
    -0.59
     Man
    -0.58
     Ar
    -0.57
    <eos>
    -0.57
     H
    -0.57
    ,
    -0.57
    POSITIVE LOGITS
    IGraphics
    1.11
    المناصب
    1.07
    CloseOperation
    1.06
    存于互联网档案馆
    1.04
    tagHelperRunner
    1.04
    parsedMessage
    1.02
    adaptiveStyles
    1.00
    ✨:
    0.99
     فريبيس
    0.99
    SharedCtor
    0.99
    Activations Density 0.016%

    No Known Activations