© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 113941
    Prev
    Next
    INDEX
    Explanations

    serialization data or news headings

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

    The neuron fires on standalone numeric literals (e.g. numbers in code or JSON), detecting number tokens in structured/textual data.

    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
     și
    -2.34
    ristiano
    -2.30
    瑨
    -2.28
    湉
    -2.23
     και
    -2.23
    folium
    -2.22
     organizador
    -2.16
     pilih
    -2.14
    羋
    -2.13
     but
    -2.11
    POSITIVE LOGITS
    ’
    3.25
    a
    3.14
            
    2.94
          
    2.61
       
    2.53
    as
    2.48
     ‘
    2.45
    =
    2.42
    '
    2.34
                  
    2.33
    Activations Density 0.001%

    No Known Activations