© 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. 69111
    Prev
    Next
    INDEX
    Explanations

    numbers in parentheses

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

    The neuron is detecting numeric expressions—especially numbers (including decimals or percentages) and counts—when they appear in or near parentheses.

    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
     because
    -2.33
     while
    -2.09
     what
    -2.05
     it
    -1.78
    what
    -1.68
    :
    -1.64
     two
    -1.63
    #
    -1.55
     What
    -1.52
     on
    -1.45
    POSITIVE LOGITS
     juzg
    1.59
    reli
    1.56
     opinie
    1.55
     praktisch
    1.52
    pij
    1.49
    鴒
    1.48
     tafel
    1.48
    ⚹
    1.48
     betek
    1.47
     melk
    1.47
    Activations Density 0.035%

    No Known Activations