© 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. 71188
    Prev
    Next
    INDEX
    Explanations

    U.S. cities and states

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

    This neuron detects the “—” (double‐dash) punctuation marker (and its accompanying comma/period) used to separate the dateline (e.g. “CITY, STATE. —”) from the article text.

    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
     traitements
    -1.88
     posteriormente
    -1.77
     rembour
    -1.75
     véhic
    -1.73
     kontinuier
    -1.64
     débats
    -1.63
     secado
    -1.61
    所说的
    -1.61
     donnant
    -1.58
     dépenses
    -1.58
    POSITIVE LOGITS
     at
    2.42
     with
    2.05
    //
    1.73
     of
    1.72
    thodoxy
    1.70
     gustar
    1.66
     for
    1.65
     may
    1.64
     For
    1.58
    z
    1.58
    Activations Density 0.030%

    No Known Activations