© 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. 22-GEMMASCOPE-RES-131K
    7. 33218
    Prev
    Next
    INDEX
    Explanations

    sending emails

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

    The neuron activates whenever the text refers to mailing/email‐sending (e.g. tokens like “Mail,” “mail,” “send,” or “Email”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    grees
    -1.25
     ల
    -1.20
     рослин
    -1.18
     tropicales
    -1.08
     saks
    -1.08
     /
    -1.06
     kompetisi
    -1.00
     растений
    -1.00
    fil
    -0.98
    licante
    -0.98
    POSITIVE LOGITS
     send
    1.30
     sending
    1.22
     sends
    1.21
    Sending
    1.21
     sent
    1.20
     SqlParameter
    1.20
     message
    1.18
     Sending
    1.16
     envoyer
    1.16
    vgili
    1.13
    Activations Density 0.028%

    No Known Activations