© 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. 10-GEMMASCOPE-RES-131K
    7. 103906
    Prev
    Next
    INDEX
    Explanations

    during and after

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

    The neuron is broadly detecting high-frequency English function words (common stopwords like “the,” “and,” “that,” “through,” “during,” etc.).

    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
     valioso
    -2.19
     afectada
    -2.03
     cotidiano
    -2.00
     don
    -1.93
     durchaus
    -1.90
     künftig
    -1.87
    瓩
    -1.85
    ⵢ
    -1.83
     finds
    -1.82
     Anschließend
    -1.82
    POSITIVE LOGITS
    o
    2.23
    y
    2.06
     this
    2.05
     these
    1.88
     At
    1.87
     richten
    1.86
    h
    1.86
     But
    1.80
    Según
    1.80
     those
    1.77
    Activations Density 0.007%

    No Known Activations