© 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-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 40424
    Prev
    Next
    INDEX
    Explanations

    closing Punctuation

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

    The neuron responds to structural markup and section‐header cues (e.g. closing HTML/XML tags like </em>, <i> markers, “abstract:”, code include lines, etc.), i.e. it flags document‐structure or section-boundary tokens.

    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
     technisch
    -1.63
    러나
    -1.63
    agine
    -1.50
    rageous
    -1.47
    :"#
    -1.45
    etted
    -1.41
    rainian
    -1.38
     incluindo
    -1.37
    ᆼ
    -1.36
     éste
    -1.35
    POSITIVE LOGITS
     أيض
    1.42
    تصویر
    1.38
    气氛
    1.34
    蘆
    1.31
     sui
    1.30
    还有一个
    1.30
     wspania
    1.29
     sprints
    1.28
     LIABLE
    1.27
    </i>
    1.27
    Activations Density 0.007%

    No Known Activations