© 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. 4404
    Prev
    Next
    INDEX
    Explanations

    Dul or Jul prefixes

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

    The neuron strongly responds to the character sequence “dul” (e.g. in dulcet, Dulworth, duloxetine).

    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
    typec
    -0.68
     tử
    -0.67
    netz
    -0.66
     związ
    -0.66
    ;">
    -0.66
    igua
    -0.66
    Fren
    -0.65
     Oriented
    -0.65
    Groetjes
    -0.64
     dedicado
    -0.64
    POSITIVE LOGITS
     dul
    1.09
     Dul
    0.90
    тнес
    0.88
     dull
    0.88
     roar
    0.85
    TEP
    0.80
     Dull
    0.77
    expose
    0.74
    भ
    0.73
     pauvres
    0.70
    Activations Density 0.035%

    No Known Activations