© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 14879
    Prev
    Next
    INDEX
    Explanations

    possessive names before related nouns

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

    This neuron primarily detects the token “s” when it appears as a standalone marker—most often the ’s possessive suffix.

    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
     сам
    -1.48
     sowohl
    -1.38
     merely
    -1.37
    ஂ
    -1.36
    ANIUM
    -1.34
     versátil
    -1.34
    achuca
    -1.34
     acrí
    -1.31
     everything
    -1.30
     anything
    -1.27
    POSITIVE LOGITS
    )
    1.70
    ?
    1.45
     their
    1.43
     lachen
    1.41
     diyor
    1.39
     okazji
    1.35
    <td>
    1.35
     themselves
    1.34
    petits
    1.32
    ,
    1.32
    Activations Density 0.011%

    No Known Activations