© 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. 41654
    Prev
    Next
    INDEX
    Explanations

    dates and times following "on"

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

    The neuron is specialized to spike on numeric metadata tokens—especially dates, times, counts, and ID numbers found in headers or “submitted/posted by … on …” lines.

    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
     英文
    -1.37
    Lyric
    -1.31
     would
    -1.25
     颜
    -1.22
    Slik
    -1.20
    Отзывы
    -1.18
     uhr
    -1.18
    ronpa
    -1.16
     fiancée
    -1.16
    seded
    -1.16
    POSITIVE LOGITS
     ​
    1.35
     endet
    1.34
     ‏
    1.30
    on
    1.29
     Autres
    1.27
    from
    1.27
     who
    1.27
    ¨
    1.25
     adicion
    1.25
     în
    1.23
    Activations Density 0.058%

    No Known Activations