© 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. 67642
    Prev
    Next
    INDEX
    Explanations

    html tags

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

    The neuron fires strongest on the angle-bracket characters of HTML/XML‐style tags, i.e. it detects markup tag delimiters.

    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
     and
    -2.28
     now
    -2.17
     There
    -2.02
     from
    -1.65
     Don
    -1.61
     Now
    -1.55
      
    -1.54
     still
    -1.53
     as
    -1.52
     (
    -1.51
    POSITIVE LOGITS
    
    1.85
     japoneses
    1.81
    あぁ
    1.81
     vives
    1.79
    
    1.78
     désert
    1.73
    
    1.71
    莚
    1.70
     célè
    1.70
    うーん
    1.69
    Activations Density 0.008%

    No Known Activations