© 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. 21944
    Prev
    Next
    INDEX
    Explanations

    punctuation and citations

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

    The neuron is picking out tokens that are part of structured numeric or reference notation—such as numbers, decimal points, brackets/parentheses around citations or measurements, and related punctuation.

    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
     concorda
    -1.68
     senin
    -1.61
     diki
    -1.59
    你是不是
    -1.55
     Lunes
    -1.52
     Resmi
    -1.52
     seksi
    -1.50
    其实
    -1.50
    Ще
    -1.49
     camada
    -1.48
    POSITIVE LOGITS
     on
    2.03
     from
    2.02
     as
    1.90
     most
    1.89
     to
    1.81
     me
    1.69
     very
    1.64
     of
    1.60
     making
    1.59
     without
    1.58
    Activations Density 0.293%

    No Known Activations