© 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. 110690
    Prev
    Next
    INDEX
    Explanations

    negations and contractions

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

    The neuron detects negative contractions (the “n't” suffix in words like don’t, aren’t, didn’t, wasn’t).

    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
    ﮨ
    -2.13
    訁
    -1.95
    teten
    -1.95
     مقالات
    -1.76
     didik
    -1.72
    谖
    -1.71
    cripts
    -1.69
    〼
    -1.68
     harten
    -1.68
    ほんと
    -1.68
    POSITIVE LOGITS
     doesn
    2.11
     don
    1.89
     have
    1.87
     aren
    1.77
    ),
    1.67
     tell
    1.66
     haven
    1.66
     didn
    1.59
    le
    1.57
    in
    1.56
    Activations Density 0.073%

    No Known Activations