© 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. 2954
    Prev
    Next
    INDEX
    Explanations

    mentions of formal actions, plans or agreements, and how they were set up and handled.

    oai_token-act-pair · gemini-2.0-flash

    draw

    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
    ksz
    -0.61
    sium
    -0.55
     sulit
    -0.51
     cured
    -0.51
    ianum
    -0.50
     filtrate
    -0.49
    obox
    -0.48
    AKI
    -0.47
     palais
    -0.47
    erapeu
    -0.47
    POSITIVE LOGITS
     draw
    1.26
     drawn
    1.19
     drew
    1.14
     draws
    1.13
    draw
    1.06
     drawing
    1.05
     Draw
    1.03
     Draws
    1.02
     Drawing
    1.02
     Drawn
    1.00
    Activations Density 0.806%

    No Known Activations