© 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-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 38794
    Prev
    Next
    INDEX
    Explanations

    fieldset legend or specific technical terms

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron fires on HTML tags that group or section off content—especially form grouping tags like `<fieldset>` and its `<legend>` (and similarly structural tags such as `<blockquote>`).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ᠂
    -2.55
    齙
    -2.41
    饜
    -2.41
    躝
    -2.33
    却被
    -2.31
    蟶
    -2.31
    鼉
    -2.27
    e
    -2.23
    柃
    -2.11
    ︕
    -2.09
    POSITIVE LOGITS
    ,“
    2.63
    <bos>
    2.55
    “
    2.52
     gesamte
    2.44
     zahlreichen
    2.31
     flurry
    2.28
    '
    2.25
            
    2.25
     này
    2.19
     дополнительных
    2.17
    Activations Density 0.001%

    No Known Activations