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. 4-GEMMASCOPE-MLP-65K
    6. 45
    Prev
    Next
    INDEX
    Explanations

    references to levels of deprivation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 4-gemmascope-mlp-65k
    Configuration
    google/gemma-scope-2b-pt-mlp/layer_4/width_65k/average_l0_66
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.4.hook_mlp_out
    Hook Layer
    4
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     dep
    -2.78
     Dep
    -2.69
    dep
    -2.69
    Dep
    -2.67
     DEP
    -2.03
    DEP
    -1.91
     dép
    -1.69
     Dép
    -1.49
     deps
    -1.29
     deput
    -1.21
    POSITIVE LOGITS
    oneofs
    0.49
    CodeAttribute
    0.47
    //
    0.43
     rápidas
    0.43
     unfinished
    0.42
    elesai
    0.42
     falt
    0.41
     Adler
    0.40
     rapides
    0.40
    accepte
    0.40
    Activations Density 0.013%

    No Known Activations