© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-2B
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 6822
    Prev
    Next
    INDEX
    Explanations

    references to racial and ethnic disparities in healthcare access and experiences.

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

    information related to medical conditions and demographics, especially focusing on racial and ethnic disparities.

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    This neuron appears to detect comparisons or contrasts between racial/ethnic groups, particularly focusing on mentions of white people compared to other racial minorities. It shows high activations for phrases that juxtapose different racial groups or discuss racial demographics and disparities. However, I want to note that

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @bot

    the word "white"

    np_acts-logits-general · gemini-2.0-flash

    the word white

    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
    Jeografia
    -0.81
    Tikang
    -0.64
     CreateTagHelper
    -0.63
     fubject
    -0.60
     femininas
    -0.59
    featureID
    -0.58
     kysy
    -0.58
     ſeveral
    -0.57
     sogget
    -0.57
    ohjel
    -0.57
    POSITIVE LOGITS
     white
    2.15
    white
    1.85
     White
    1.81
    White
    1.72
     WHITE
    1.66
     whites
    1.50
    WHITE
    1.45
    白
    1.40
     Whites
    1.31
    whites
    1.28
    Activations Density 0.336%

    No Known Activations