© 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. 127558
    Prev
    Next
    INDEX
    Explanations

    phrases with key nouns and following words

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

    The neuron fires on standalone capitalized nouns and short phrases used as section headings or list‐item titles (e.g. menu or bullet labels).

    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
     to
    -2.17
    quite
    -1.85
     ۱۵
    -1.68
    however
    -1.68
    ⩕
    -1.66
     trône
    -1.64
     éton
    -1.63
     بالإضافة
    -1.63
     โดย
    -1.61
     จะ
    -1.61
    POSITIVE LOGITS
     first
    2.08
     nine
    2.06
     four
    1.98
     after
    1.91
     when
    1.84
     sixteen
    1.84
     thirteen
    1.84
     seven
    1.81
     adequate
    1.80
     sooooo
    1.79
    Activations Density 0.121%

    No Known Activations