© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 8235
    Prev
    Next
    INDEX
    Explanations

    address/location format

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

    The neuron strongly activates on phrases that introduce or list email‐related contact information (e.g. “email addresses,” “possible email formats for…”).

    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_24_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
     musste
    0.74
    реди
    0.70
     земе
    0.70
     conoscere
    0.69
    łeś
    0.68
    ඬ
    0.68
    igungs
    0.67
     Commonly
    0.67
     ricordare
    0.67
    hrung
    0.66
    POSITIVE LOGITS
     tenets
    0.78
    톱
    0.68
     tasked
    0.67
     aimed
    0.66
     thereby
    0.66
    
    0.65
    
    0.65
     extol
    0.64
     albeit
    0.64
    これにより
    0.64
    Activations Density 0.007%

    No Known Activations