© 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. 22-GEMMASCOPE-TRANSCODER-16K
    7. 7148
    Prev
    Next
    INDEX
    Explanations

    a high frequency of short, common words, used to link larger components of a sentence

    oai_token-act-pair · gemini-2.0-flash

    utmost importance

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
     itſelf
    -1.00
     Efq
    -1.00
     Monfieur
    -0.99
     Theſe
    -0.98
     myſelf
    -0.98
     Wikimedijinoj
    -0.96
     Jefus
    -0.95
     iſt
    -0.91
     $_"
    -0.90
     ―――――
    -0.90
    POSITIVE LOGITS
     full
    0.52
     up
    0.51
     ut
    0.50
     bit
    0.46
    up
    0.44
      
    0.44
     ri
    0.42
     et
    0.41
     pri
    0.41
    gi
    0.41
    Activations Density 1.263%

    No Known Activations