© 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. 14158
    Prev
    Next
    INDEX
    Explanations

    words and phrases related to professional, political, or technical language

    oai_token-act-pair · gemini-2.0-flash

    business, services, economics

    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
    LookAnd
    -1.27
    SourceChecksum
    -1.16
     resourceCulture
    -1.08
    parsedMessage
    -1.04
     &___
    -1.04
    WithIOException
    -1.02
     wireType
    -1.02
    AndEndTag
    -1.00
    oneofs
    -1.00
    mergeFrom
    -0.99
    POSITIVE LOGITS
    quiel
    0.58
    樣的
    0.52
    ITIONAL
    0.48
    echt
    0.47
    Dowell
    0.46
    和我
    0.45
    gebraucht
    0.45
    和小
    0.44
    keits
    0.44
    י
    0.42
    Activations Density 5.780%

    No Known Activations