© 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. 21-GEMMASCOPE-TRANSCODER-16K
    7. 10676
    Prev
    Next
    INDEX
    Explanations

    words used as light verbs, such as "has", "have', "do", "gives", "is", "turned", "came", and "had" when they are followed by complements

    oai_token-act-pair · gemini-2.0-flash

    expressions of nothingness or absence

    np_token-act-pair-logits · gpt-4o-mini

    nothing, zero

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.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
    MigrationBuilder
    -0.81
     saites
    -0.71
     AssemblyProduct
    -0.69
     ExecuteAsync
    -0.68
     تانيه
    -0.66
     '\\;'
    -0.64
    دانشنامهٔ
    -0.59
    ContentAsync
    -0.58
     ostavi
    -0.56
     demografica
    -0.54
    POSITIVE LOGITS
     zero
    1.54
     absolutely
    1.53
     nothing
    1.38
     Zero
    1.31
    absolutely
    1.28
     Absolutely
    1.27
    Absolutely
    1.26
    zero
    1.25
     ZERO
    1.22
     nil
    1.17
    Activations Density 3.075%

    No Known Activations