© 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. 95410
    Prev
    Next
    INDEX
    Explanations

    introducing definitions hereafter

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

    the neuron detects formal definition phrases that introduce or abbreviate a term (e.g. “hereafter referred to as,” “henceforth used for,” “often abbreviated as”).

    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
     still
    -1.06
     damals
    -1.06
     didn
    -1.06
     wasn
    -1.00
     had
    -1.00
     couldn
    -0.98
     like
    -0.97
    دانلود
    -0.96
    مقدمه
    -0.94
     damaligen
    -0.93
    POSITIVE LOGITS
     hereinafter
    1.23
     henceforth
    1.21
     всички
    1.21
     všet
    1.13
     budeme
    1.13
     necesarios
    1.12
     hereafter
    1.09
     afbeeldingen
    1.09
     will
    1.09
     tých
    1.09
    Activations Density 0.052%

    No Known Activations