© 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. 98943
    Prev
    Next
    INDEX
    Explanations

    the entity name

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

    The neuron is picking out the leading words that start formal section headings or titles (e.g. “The,” “This,” “In,” etc.) at the beginning of new document sections.

    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
     there
    -1.96
     of
    -1.65
     if
    -1.60
    さんへ
    -1.54
     حتی
    -1.53
    可能有
    -1.52
     Pokud
    -1.43
    
    -1.38
     Apakah
    -1.38
     nadie
    -1.37
    POSITIVE LOGITS
    ↵↵
    1.61
     juf
    1.53
     latter
    1.50
     "")
    1.48
    lasyon
    1.41
    ETERS
    1.38
     last
    1.38
    álu
    1.37
    ℳ
    1.35
     '-':
    1.34
    Activations Density 0.217%

    No Known Activations