© 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. 21123
    Prev
    Next
    INDEX
    Explanations

    registered office or seat

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

    The neuron fires on tokens in formal “registered office” or similar corporate/legal registration blocks, flagging numeric and address fields (e.g. “Registered office is …”).

    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
    </b>
    -1.46
    こい
    -1.38
     kemeja
    -1.30
    ти
    -1.29
     We
    -1.28
    ва
    -1.27
     上午
    -1.26
    »,
    -1.23
     「
    -1.23
     два
    -1.21
    POSITIVE LOGITS
     headquarters
    1.47
     discernible
    1.22
    蛻
    1.21
     גם
    1.20
    škai
    1.19
     characteristics
    1.16
     residences
    1.16
    年齡
    1.14
    éviter
    1.13
     desafíos
    1.13
    Activations Density 0.013%

    No Known Activations