© 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. 99765
    Prev
    Next
    INDEX
    Explanations

    U followed by numbers

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

    The neuron consistently fires on tokens that are “U” immediately followed by one or more digits—that is, letter-number identifiers like U1, U2, U3, 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
     longtime
    -1.77
    原始内容存档于
    -1.67
    鹋
    -1.63
     perdiendo
    -1.60
     apaixon
    -1.56
    Cdb
    -1.56
     surpresa
    -1.52
    '
    -1.48
     svůj
    -1.48
     ziemlich
    -1.46
    POSITIVE LOGITS
     of
    2.03
    neling
    1.55
    the
    1.48
    gencias
    1.47
     or
    1.45
    ,
    1.44
    кові
    1.39
    »
    1.39
    ื่อน
    1.38
    “
    1.34
    Activations Density 0.032%

    No Known Activations