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. 1200
    Prev
    Next
    INDEX
    Explanations

    formal legal language and judicial terminology

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

    components related to official documentation, particularly dates and names

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

    legal citations and judgments

    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
     nahilalakip
    -0.62
    SourceChecksum
    -0.61
    SharedDtor
    -0.58
    Gambas
    -0.57
     Administrativna
    -0.56
    Manbalar
    -0.56
    Autoritní
    -0.54
    TRIBUN
    -0.54
     }{@
    -0.52
    jooq
    -0.52
    POSITIVE LOGITS
    FieldDescriptor
    0.48
    ждую
    0.47
    RUnlock
    0.47
     parseFrom
    0.45
    schedulers
    0.45
     horaires
    0.45
    IsMutable
    0.44
     chrétienne
    0.44
    pite
    0.43
     CYTHON
    0.43
    Activations Density 0.055%

    No Known Activations