© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 21-GEMMASCOPE-TRANSCODER-16K
    7. 1858
    Prev
    Next
    INDEX
    Explanations

    mentions of groups of people from the past

    oai_token-act-pair · gemini-2.0-flash

    forms of the verb "to be"

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

    say "was" or "were"

    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
    참고
    -0.53
    hésite
    -0.50
    FunctionFlags
    -0.47
     is
    -0.45
    ciate
    -0.44
    лении
    -0.43
     Public
    -0.42
    textTheme
    -0.42
     do
    -0.41
    Präsident
    -0.41
    POSITIVE LOGITS
     was
    1.11
    Was
    1.09
     Was
    1.04
     WAS
    0.93
    was
    0.92
     were
    0.79
    Wasn
    0.75
     Foi
    0.66
     Wasn
    0.66
    Fue
    0.65
    Activations Density 1.292%

    No Known Activations