© 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. 7260
    Prev
    Next
    INDEX
    Explanations

    words relating to formal agreements, politics, and official institutions.

    oai_token-act-pair · gemini-2.0-flash

    legal

    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
    ValueGeneration
    -0.63
    enderror
    -0.63
    XtraReports
    -0.59
     FormData
    -0.51
     ägg
    -0.50
    arbeitung
    -0.48
     Hayat
    -0.48
     GenerationType
    -0.47
     CreateTagHelper
    -0.47
     Hibernate
    -0.47
    POSITIVE LOGITS
     legal
    2.28
    legal
    2.00
     Legal
    1.95
    Legal
    1.91
     legally
    1.90
     LEGAL
    1.89
     lawful
    1.77
    LEGAL
    1.70
     legality
    1.70
     legale
    1.63
    Activations Density 4.076%

    No Known Activations