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. 10894
    Prev
    Next
    INDEX
    Explanations

    descriptions of places and their characteristics

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

    Adjective intensifier ("very," "bit")

    np_acts-logits-general · gemini-2.0-flash

    cleanliness

    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
    tvguidetime
    -0.77
     AssemblyCulture
    -0.71
    AddTagHelper
    -0.63
     &___
    -0.58
    Personendaten
    -0.58
    <bos>
    -0.57
    MessageTagHelper
    -0.56
     noDo
    -0.56
     виправивши
    -0.56
    юз
    -0.55
    POSITIVE LOGITS
     clean
    0.89
     spacious
    0.83
     cleanliness
    0.76
     Clean
    0.72
     accogli
    0.68
     bersih
    0.67
     cleanest
    0.67
    clean
    0.66
    Clean
    0.65
     CLEAN
    0.64
    Activations Density 0.079%

    No Known Activations