© 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. 22-GEMMASCOPE-RES-131K
    7. 130323
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on “stands for” (and similar) when introducing or defining an acronym.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    acronyms and abbreviations explained

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    .
    -1.10
     then
    -1.05
     visit
    -1.04
     
    -1.02
    Then
    -0.98
     recent
    -0.97
    觉得
    -0.96
     re
    -0.94
     have
    -0.94
    Because
    -0.94
    POSITIVE LOGITS
     acronym
    1.52
     abbreviation
    1.41
     abbreviations
    1.23
     initials
    1.23
     pantal
    1.16
     serre
    1.16
     acronyms
    1.15
     ۱۷
    1.15
     ۲۵
    1.13
    什么呢
    1.13
    Activations Density 0.030%

    No Known Activations