© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 22-GEMMASCOPE-TRANSCODER-16K
    7. 15883
    Prev
    Next
    INDEX
    Explanations

    emotional punctuation in various contexts, including code, birth stories, and blog posts.

    oai_token-act-pair · gemini-2.0-flash

    so

    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
     therefore
    -0.71
     derfor
    -0.56
     portanto
    -0.54
    เลย
    -0.51
     unknownFields
    -0.51
     Therefore
    -0.49
     pues
    -0.49
     daher
    -0.48
    therefore
    -0.48
    Therefore
    -0.47
    POSITIVE LOGITS
     so
    3.05
    so
    2.23
     så
    1.80
    So
    1.69
     così
    1.67
     так
    1.63
     SO
    1.55
    如此
    1.48
     begitu
    1.46
     So
    1.45
    Activations Density 3.650%

    No Known Activations