© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 105169
    Prev
    Next
    INDEX
    Explanations

    Austral, Uttar Pradesh, numerals, unfor

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

    The neuron fires strongly on uncommon, content-heavy subword tokens—that is, rare or technical vocabulary rather than common function words.

    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
     literaria
    -2.81
     IBRARY
    -2.67
     leaſt
    -2.64
    ðsíða
    -2.48
    ÇÃO
    -2.48
    OUTUBE
    -2.33
    𞤢
    -2.30
    샅
    -2.30
    ⓦ
    -2.27
    。"
    -2.25
    POSITIVE LOGITS
    是他
    2.61
    u
    2.42
    ăzi
    2.41
    しないで
    2.41
    釹
    2.39
    糌
    2.33
    
    2.25
     sadece
    2.25
    桕
    2.23
     siguran
    2.20
    Activations Density 0.019%

    No Known Activations