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. 21-GEMMASCOPE-RES-16K
    6. 2976
    Prev
    Next
    INDEX
    Explanations

    references to data collection and analysis processes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.hook_resid_post
    Hook Layer
    21
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    TagHelper
    -0.65
    Named
    -0.57
     Named
    -0.51
     transfieras
    -0.51
     /\.(
    -0.51
     mpl
    -0.50
    wäh
    -0.50
     Crocker
    -0.48
     Magdalen
    -0.48
    Featured
    -0.47
    POSITIVE LOGITS
     collected
    1.97
     gathered
    1.70
    collected
    1.65
     collecting
    1.61
     collect
    1.60
     gathering
    1.55
     collects
    1.51
     Collecting
    1.50
     Collected
    1.49
     Collect
    1.44
    Activations Density 0.632%

    No Known Activations