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. 3030
    Prev
    Next
    INDEX
    Explanations

    references to specific locations, particularly Puerto Rico and its associated entities

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

    Locations including "Puerto" or "Rico."

    np_acts-logits-general · gemini-2.0-flash

    Puerto Rico and places

    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
    IntoConstraints
    -0.78
    tagHelper
    -0.73
    httphttps
    -0.57
    AnchorStyles
    -0.56
    __*/
    -0.55
     SwitchCompat
    -0.55
     Chilli
    -0.54
     שוליים
    -0.54
     McKinnon
    -0.54
    ciatore
    -0.53
    POSITIVE LOGITS
     Puerto
    0.98
    Puerto
    0.88
     Guam
    0.82
     puerto
    0.61
     Ponce
    0.53
     Rican
    0.52
     Calvo
    0.52
     onPostExecute
    0.50
    puerto
    0.50
    contentLoaded
    0.49
    Activations Density 0.222%

    No Known Activations