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. 978
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal proceedings and court opinions

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

    legal terms and processes related to court cases

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

    Italy and technical terms

    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
    queeze
    -0.53
    fürst
    -0.49
     DOMAIN
    -0.48
     universale
    -0.48
    aticano
    -0.48
     gypsum
    -0.46
    Squares
    -0.46
     Amalia
    -0.46
     Sulfate
    -0.45
    ordnet
    -0.45
    POSITIVE LOGITS
    SourceChecksum
    0.70
    TintMode
    0.69
     transfieras
    0.67
     nahilalakip
    0.66
    EndInit
    0.60
    FunctionFlags
    0.55
    adpleegd
    0.55
    AddTagHelper
    0.54
    InvalidProtocol
    0.54
    usitis
    0.54
    Activations Density 0.060%

    No Known Activations