© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 12661
    Prev
    Next
    INDEX
    Explanations

    names and emails

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

    The neuron activates on the username parts (the local‐part) of email addresses.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    gmail
    0.62
    ̶
    0.61
    @
    0.61
     infrastructures
    0.58
    infrastructure
    0.57
     Goodreads
    0.56
     pinterest
    0.55
     gmail
    0.54
     infrastructure
    0.54
    -
    0.54
    POSITIVE LOGITS
     irritation
    0.57
     प्रारम्भ
    0.56
    fbox
    0.55
    слары
    0.54
    imethyl
    0.54
     dernier
    0.54
     yapt
    0.54
     isomers
    0.54
    ییر
    0.54
    przy
    0.54
    Activations Density 0.009%

    No Known Activations