© 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. 83839
    Prev
    Next
    INDEX
    Explanations

    GPL license

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

    The neuron fires on isolated content‐heavy tokens—proper names or technical keywords and acronyms—essentially flagging named entities and domain‐specific terms.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     We
    -3.59
    o
    -3.41
    l
    -3.38
    =
    -2.80
    a
    -2.80
    t
    -2.53
    e
    -2.44
     -
    -2.41
    f
    -2.36
     siendo
    -2.28
    POSITIVE LOGITS
    鬭
    3.20
    嬋
    2.73
    
    2.72
    ⡯
    2.67
    い
    2.61
    گوشی
    2.61
    
    2.58
    臵
    2.56
    ዡ
    2.53
    怎么回事
    2.50
    Activations Density 0.029%

    No Known Activations