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. 2399
    Prev
    Next
    INDEX
    Explanations

    phrases related to terms of office and duration in a political context

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

    length of time served

    np_acts-logits-general · gemini-2.0-flash

    term lengths

    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
    migrationBuilder
    -0.73
    endphp
    -0.64
     okuyayım
    -0.63
     فريبيس
    -0.58
    AccessorTable
    -0.57
    [][]
    -0.57
    rago
    -0.55
    Referanser
    -0.55
     disambiguazione
    -0.55
     nakalista
    -0.54
    POSITIVE LOGITS
     term
    1.03
    Term
    0.94
     Term
    0.92
    term
    0.86
     TERM
    0.83
    TERM
    0.76
     five
    0.70
     four
    0.68
    terms
    0.66
     six
    0.64
    Activations Density 0.257%

    No Known Activations