© 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. 130160
    Prev
    Next
    INDEX
    Explanations

    The neuron selectively fires on isolated acronym- or code-style tokens (e.g. technical abbreviations, gene/enzyme names, month names written as standalone capitalized words, and similar alphanumeric identifiers).

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

    specific identifiers and codes

    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
    Má
    -1.00
     which
    -0.91
    这是一
    -0.90
    嬬
    -0.89
    錨
    -0.89
    日至
    -0.88
      “
    -0.87
     now
    -0.85
     their
    -0.84
    Transcripción
    -0.83
    POSITIVE LOGITS
    therefore
    1.02
      
    1.01
    ほんとに
    1.00
    Therefore
    0.99
    much
    0.99
     ,'
    0.96
    шёл
    0.96
     
    0.92
    咂
    0.90
     kwamba
    0.89
    Activations Density 0.105%

    No Known Activations