© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 35896
    Prev
    Next
    INDEX
    Explanations

    UI code structure

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

    The neuron is activated by HTML/XML tag syntax (e.g. the `<`, tag names, and attributes).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     gelungen
    -1.16
     rémun
    -1.16
     mehreren
    -1.09
    zahlt
    -1.09
     fréquent
    -1.09
     témo
    -1.08
     douze
    -1.07
     célébr
    -1.07
     rembourse
    -1.05
     contrô
    -1.04
    POSITIVE LOGITS
     with
    1.37
     mainly
    1.13
     continues
    1.09
     still
    1.00
    The
    0.96
     actions
    0.96
     uses
    0.95
     then
    0.95
     before
    0.94
     THE
    0.94
    Activations Density 0.006%

    No Known Activations