© 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. 84648
    Prev
    Next
    INDEX
    Explanations

    codesemicolon

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

    The neuron activates on numeric tokens (e.g. version numbers or numeric literals) in code.

    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
      
    -1.77
     "
    -1.77
     don
    -1.73
    -
    -1.65
    1
    -1.53
       
    -1.52
     You
    -1.51
     weren
    -1.51
     целый
    -1.48
     специальные
    -1.45
    POSITIVE LOGITS
     árbitro
    1.71
     sexto
    1.69
     arus
    1.63
     anún
    1.61
     benda
    1.59
     kirim
    1.59
     länder
    1.59
    بیشتر
    1.57
    erintah
    1.57
     kini
    1.56
    Activations Density 0.011%

    No Known Activations