© 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. 2897
    Prev
    Next
    INDEX
    Explanations

    words relating to people and dates, particularly in a sports context

    oai_token-act-pair · gemini-2.0-flash

    Sports players and coaches

    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
    +#+#
    -0.92
     kasarigan
    -0.55
    μιουργ
    -0.49
    ьаж
    -0.47
     team
    -0.47
     home
    -0.44
    permitAll
    -0.42
    οντας
    -0.41
    ApiController
    -0.41
    홈
    -0.40
    POSITIVE LOGITS
     AssemblyProduct
    0.65
    ","");
    0.60
     coached
    0.60
    Abp
    0.59
    '][]
    0.59
    mergeFrom
    0.58
    DIPSETTING
    0.57
     giocare
    0.55
     vincere
    0.54
     Playing
    0.53
    Activations Density 2.937%

    No Known Activations