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 - 16k
    5. 15-GEMMASCOPE-RES-16K
    6. 2401
    Prev
    Next
    INDEX
    Explanations

    instances of numbers and their associations within the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     houſe
    -0.73
     longtemps
    -0.69
     purpoſe
    -0.67
     chrétiens
    -0.67
    ſelf
    -0.65
     itſelf
    -0.65
     pleaſure
    -0.63
     quelcon
    -0.63
     traités
    -0.62
     normaux
    -0.62
    POSITIVE LOGITS
     three
    0.93
     two
    0.84
     different
    0.83
     separate
    0.83
     four
    0.82
     dozen
    0.81
     distinct
    0.71
     five
    0.69
     more
    0.69
     successive
    0.68
    Activations Density 0.554%

    No Known Activations