© 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. 10803
    Prev
    Next
    INDEX
    Explanations

    words frequently found in news articles and magazines, particularly those related to publications and journalism.

    oai_token-act-pair · gemini-2.0-flash

    articles in other languages

    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
     Theſe
    -0.77
     Anſ
    -0.72
     Majefty
    -0.70
     Inſ
    -0.70
     Beſ
    -0.69
    ſelf
    -0.66
     RIPRODUZIONE
    -0.62
     Reſ
    -0.62
     pleaſure
    -0.59
     Houſe
    -0.57
    POSITIVE LOGITS
     La
    1.34
     Les
    1.25
    La
    1.19
     Le
    1.16
    Les
    1.11
     El
    1.07
     la
    0.96
     Das
    0.95
    Le
    0.93
     Il
    0.93
    Activations Density 2.694%

    No Known Activations