© 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. 23-GEMMASCOPE-TRANSCODER-16K
    7. 14068
    Prev
    Next
    INDEX
    Explanations

    a variety of different things, but it seems to activate more often on variables, numbers and letters within mathematical equations or code

    oai_token-act-pair · gemini-2.0-flash

    code, formulas

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_23/width_16k/average_l0_25
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.23.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
     P
    -1.15
     S
    -1.12
     D
    -1.03
     F
    -1.02
     C
    -1.02
     T
    -1.02
     G
    -1.01
     E
    -0.99
     M
    -0.98
     B
    -0.96
    POSITIVE LOGITS
     betweenstory
    0.59
     समीक्षक
    0.54
     lisse
    0.54
     AdapterView
    0.54
     q
    0.53
    ianum
    0.52
     ТАК
    0.52
    makeConstraints
    0.49
    ΤΗ
    0.48
     r
    0.48
    Activations Density 5.590%

    No Known Activations