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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 21-GEMMASCOPE-TRANSCODER-16K
    7. 9920
    Prev
    Next
    INDEX
    Explanations

    words that were associated with how slaves were "well-treated" and how people of color and other groups were/are manipulated.

    oai_token-act-pair · gemini-2.0-flash

    forms of the verb "be"

    np_token-act-pair-logits · gpt-4o-mini

    and,

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     their
    -2.55
    Their
    -2.45
    their
    -2.30
     Their
    -2.23
     THEIR
    -1.87
     kanilang
    -1.82
    他们的
    -1.76
     leur
    -1.73
    彼らの
    -1.70
    他們的
    -1.69
    POSITIVE LOGITS
     are
    0.66
     were
    0.59
    ratto
    0.57
    lujah
    0.52
    are
    0.52
     aren
    0.51
     rowspan
    0.50
     يتيمه
    0.49
     हैं
    0.49
     ARE
    0.49
    Activations Density 5.700%

    No Known Activations