© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 32879
    Prev
    Next
    INDEX
    Explanations

    XML and configuration tags

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

    The neuron fires on the start of XML/HTML closing tags (the “</” sequence).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     their
    -1.60
     those
    -1.45
     more
    -1.39
     on
    -1.23
     all
    -1.21
     just
    -1.19
     quite
    -1.19
     some
    -1.19
     what
    -1.16
     several
    -1.16
    POSITIVE LOGITS
     trám
    1.45
     カットソー
    1.30
    Ingredienti
    1.27
    킁
    1.25
     パッド
    1.21
    throughout
    1.20
    förmig
    1.20
     modelAndView
    1.19
     —
    1.19
     放題
    1.14
    Activations Density 0.002%

    No Known Activations