© 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. 10765
    Prev
    Next
    INDEX
    Explanations

    Japanese and Chinese particles/suffixes

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

    This neuron fires on genitive/possessive particles in Asian languages (e.g. Japanese の or Chinese 的) that link one noun to another.

    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
    eabouts
    -1.87
    hiva
    -1.84
    nyard
    -1.80
    antism
    -1.75
    fordshire
    -1.73
    estries
    -1.67
    neling
    -1.67
    aneity
    -1.65
    }))
    -1.65
    );
    -1.64
    POSITIVE LOGITS
    糴
    1.80
    </em>
    1.73
     ultimate
    1.71
     Vergnügen
    1.70
     own
    1.70
     purposes
    1.69
     portatil
    1.69
    橚
    1.68
     preços
    1.67
     établir
    1.66
    Activations Density 0.007%

    No Known Activations