© 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. 22-GEMMASCOPE-RES-131K
    7. 12871
    Prev
    Next
    INDEX
    Explanations

    dates with slashes or dots

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

    The neuron strongly activates on numeric tokens—especially those forming dates, times, or other multi‐digit data (e.g. “14/11/2402,” “08.09.2020,” “17/10/2014”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     wysokość
    -1.57
     what
    -1.44
     that
    -1.39
     different
    -1.38
     With
    -1.30
     \
    -1.29
     distinct
    -1.27
     waaronder
    -1.13
     długość
    -1.10
     will
    -1.09
    POSITIVE LOGITS
     coucher
    1.48
     perpé
    1.44
     bonheur
    1.43
    Kelebihan
    1.43
    ebly
    1.40
     diriger
    1.39
     plafon
    1.39
     listes
    1.39
     réjou
    1.37
     précédent
    1.36
    Activations Density 0.006%

    No Known Activations