© 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
    2. Gemma-3-27B-IT
    3. 31-GEMMASCOPE-2-TRANSCODER-262K
    4. 75497
    Prev
    Next
    INDEX
    Explanations

    The neuron appears to be sensitive to or predictive of **animation parameters in code and styling**.This is indicated by the presence of terms like `duration`, `animation`, `linear`, `repeatForever`, `autoreverses`, `easing`, `response`, `dampingFraction`, and CSS properties related to animation, alongside common code tokens like parentheses and numbers. The `TOP_POSITIVE_LOGITS` also contain terms like `xml` and `db` which are often found in structured data or code contexts. The `MAX_ACTIVATING_TOKENS` and `TOKENS_AFTER_MAX_ACTIVATING_TOKEN` lists seem to contain specific numerical or character sequences that often accompany such code snippets, serving as delimiters or values within them.animation parameters in code and styling

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_31_width_262k_l0_small_affine
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Deep
    0.29
    <0x9C>
    0.29
    deep
    0.28
     Моло
    0.28
     Lie
    0.28
     कट
    0.28
     wrinkles
    0.28
    Hi
    0.28
     landmarks
    0.27
    Moles
    0.27
    POSITIVE LOGITS
     xml
    0.33
    富有
    0.29
     db
    0.27
     対象
    0.27
     আচ্ছ
    0.27
    xml
    0.27
     Adu
    0.27
     متعدد
    0.26
     тексти
    0.26
    计算机
    0.26
    Activations Density 0.002%

    No Known Activations