© 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. 3773
    Prev
    Next
    INDEX
    Explanations

    determiner followed by noun or verb

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

    This neuron activates on words that begin new sentences—i.e. sentence-initial (capitalized) tokens.

    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
     to
    -2.30
     and
    -2.16
     first
    -2.11
     three
    -2.09
     two
    -2.06
     four
    -2.05
     can
    -1.98
     During
    -1.97
    When
    -1.97
     through
    -1.96
    POSITIVE LOGITS
    揍
    1.82
     quoc
    1.80
     of
    1.76
     ſeveral
    1.70
    basically
    1.68
    就去
    1.68
    事です
    1.65
     Dinas
    1.63
     représenter
    1.59
     SPECIFIC
    1.59
    Activations Density 0.471%

    No Known Activations