Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Gemma-2-2B
    3. 20-GEMMASCOPE-TRANSCODER-16K
    4. 14160
    Prev
    Next
    INDEX
    Explanations

    text that is racist or condescending towards developing nations and African people.

    oai_token-act-pair · gemini-2.0-flash

    less developed countries

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope/layer_20/width_16k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    yntaxException
    -0.69
    routeProvider
    -0.68
    URLException
    -0.66
    HostException
    -0.66
    Демографія
    -0.65
    EndContext
    -0.63
    StandardCharsets
    -0.63
    ArgsConstructor
    -0.62
    SourceChecksum
    -0.61
     estekak
    -0.60
    POSITIVE LOGITS
     underdeveloped
    0.78
     impoverished
    0.75
     backward
    0.75
     poor
    0.70
     undeveloped
    0.69
     poverty
    0.68
     poorer
    0.65
    Third
    0.65
     developing
    0.64
    backward
    0.63
    Activations Density 2.337%

    No Known Activations