© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 40784
    Prev
    Next
    INDEX
    Explanations

    contractions followed by verbs

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

    The neuron fires on English clitic suffixes (’s, ’m, n’t, ’ve, etc.) and similar informal filler words, i.e. contracted forms.

    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
     a
    -1.73
     introduces
    -1.71
     such
    -1.65
     is
    -1.54
     they
    -1.53
     will
    -1.45
     continue
    -1.39
    $
    -1.38
     the
    -1.38
     both
    -1.35
    POSITIVE LOGITS
    '
    1.77
    ᥬ
    1.73
     partielle
    1.66
    rocław
    1.62
     étroite
    1.47
    δας
    1.45
    theid
    1.43
    kultur
    1.41
    ！"
    1.39
    瘩
    1.38
    Activations Density 0.057%

    No Known Activations