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. 20-GEMMASCOPE-RES-16K
    6. 10130
    Prev
    Next
    INDEX
    Explanations

    references to childbirth and family milestones

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

    children or family members

    np_acts-logits-general · gemini-2.0-flash

    children and family members

    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-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    EndInit
    -0.58
     InputDecoration
    -0.56
     linkovi
    -0.53
     psycopg
    -0.53
    enment
    -0.49
    postsleuth
    -0.47
     Injectable
    -0.45
     onOptions
    -0.45
     arkadaş
    -0.45
    таратура
    -0.44
    POSITIVE LOGITS
     twins
    0.95
    AccessorTable
    0.83
     triplets
    0.76
     born
    0.75
     boys
    0.74
     boy
    0.72
     daughters
    0.69
     youngest
    0.66
     eldest
    0.66
     adorable
    0.66
    Activations Density 0.234%

    No Known Activations