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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 22-GEMMASCOPE-TRANSCODER-16K
    7. 9976
    Prev
    Next
    INDEX
    Explanations

    text from academic papers and/or sports reports/statistics

    oai_token-act-pair · gemini-2.0-flash

    once and twice

    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
    DockStyle
    -0.81
     متعلقه
    -0.71
    FormTagHelper
    -0.70
     EconPapers
    -0.69
    writeFieldEnd
    -0.64
    glie
    -0.62
    AndEndTag
    -0.59
     дописавши
    -0.59
     HttpNotFound
    -0.58
     AssemblyCulture
    -0.57
    POSITIVE LOGITS
     once
    2.11
     twice
    1.91
    once
    1.90
    twice
    1.75
    Once
    1.73
     Once
    1.72
     ONCE
    1.64
    Twice
    1.59
    ONCE
    1.56
     Twice
    1.55
    Activations Density 3.678%

    No Known Activations