© 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. 58423
    Prev
    Next
    INDEX
    Explanations

    email addresses and mail

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

    The neuron fires on occurrences of email‐related tokens (like “email,” “address,” “id,” or “@”), i.e. it detects mentions of email addresses.

    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
     not
    -1.66
     about
    -1.49
    </sup>
    -1.34
    ︷
    -1.32
     MÁS
    -1.32
     milhões
    -1.30
     vasi
    -1.30
     quase
    -1.30
     parati
    -1.29
    </b>
    -1.27
    POSITIVE LOGITS
     oleju
    1.48
    ností
    1.42
    gaged
    1.37
    犧
    1.34
    ингл
    1.27
    逕
    1.26
    雖
    1.25
    駭
    1.25
    刪
    1.21
    そのような
    1.20
    Activations Density 0.057%

    No Known Activations