© 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. 3607
    Prev
    Next
    INDEX
    Explanations

    words related to antennas and signal processing especially phase and frequency

    oai_token-act-pair · gemini-2.0-flash

    terms related to loss and degradation

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

    circuit properties

    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
    AndEndTag
    -1.12
    CloseOperation
    -0.94
    RegistryLite
    -0.83
    NameInMap
    -0.79
    homonymie
    -0.77
     }}"></
    -0.76
    itattu
    -0.72
    ftagPool
    -0.71
     surla
    -0.69
     Shakspeare
    -0.69
    POSITIVE LOGITS
     loss
    0.56
     Loss
    0.52
     di
    0.50
     star
    0.50
    Loss
    0.48
    loss
    0.46
     length
    0.44
     Co
    0.42
     ka
    0.42
     Ant
    0.42
    Activations Density 1.775%

    No Known Activations