© 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. 34-GEMMASCOPE-RES-131K
    7. 2191
    Prev
    Next
    INDEX
    Explanations

    positive expressions and actions

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

    The neuron fires on isolated “header” or section‐start tokens—words that appear at the beginnings of blocks (file‐header comments, section titles, paragraph starts).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     sneered
    -0.79
    tuur
    -0.79
    renderer
    -0.77
    ət
    -0.76
    dirname
    -0.73
    مساعدة
    -0.72
    verkauf
    -0.72
    personen
    -0.71
     desaf
    -0.70
    пры
    -0.70
    POSITIVE LOGITS
     smiling
    0.96
     waved
    0.90
     smiles
    0.89
     wave
    0.87
     coughed
    0.85
     praising
    0.84
     smile
    0.84
    cough
    0.83
     fond
    0.82
     cough
    0.82
    Activations Density 0.002%

    No Known Activations