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. MLP Out - 65k
    5. 0-GEMMASCOPE-MLP-65K
    6. 61107
    Prev
    Next
    INDEX
    Explanations

    mentions of "world class" or similar phrases indicating high quality or excellence

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 0-gemmascope-mlp-65k
    Configuration
    google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.0.hook_mlp_out
    Hook Layer
    0
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     class
    -0.75
    theless
    -0.56
    posób
    -0.54
     виправивши
    -0.51
     self
    -0.48
     vac
    -0.47
    ellschaft
    -0.46
     lune
    -0.46
    Datuak
    -0.45
    ymmetry
    -0.45
    POSITIVE LOGITS
    awtextra
    0.79
     conçus
    0.71
     vendus
    0.69
     fréquent
    0.68
     textStatus
    0.67
     concevoir
    0.66
     élevées
    0.64
     équip
    0.64
    addCriterion
    0.63
     fournisse
    0.62
    Activations Density 0.006%

    No Known Activations