© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 22-GEMMASCOPE-TRANSCODER-16K
    7. 12328
    Prev
    Next
    INDEX
    Explanations

    technical or scientific language, academic studies and references.

    oai_token-act-pair · gemini-2.0-flash

    academic texts

    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
    Not in Any Lists

    No Comments

    Negative Logits
     auffi
    -1.13
     auroit
    -1.12
     ainfi
    -1.11
     pleaſure
    -1.09
     faſt
    -1.09
     étoient
    -1.07
     avoient
    -1.04
     iſt
    -1.04
     feroit
    -0.97
     itſelf
    -0.97
    POSITIVE LOGITS
     in
    0.73
    ,
    0.70
     to
    0.67
     from
    0.66
     and
    0.65
     w
    0.65
     on
    0.64
     it
    0.64
    .
    0.64
     for
    0.61
    Activations Density 15.738%

    No Known Activations