© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 9580
    Prev
    Next
    INDEX
    Explanations

    US states and cities

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

    This neuron specializes in detecting numeric and numeric-like tokens (e.g. street addresses, phone numbers, dates, monetary amounts).

    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
     आंध्र
    0.81
     Andhra
    0.71
     जिलाधिकारी
    0.71
     तेलु
    0.67
    зербайд
    0.67
    婺
    0.66
    衢
    0.66
    畈
    0.66
    皖
    0.65
     Zhejiang
    0.63
    POSITIVE LOGITS
     Minnesota
    2.25
    Minnesota
    2.14
     Minneapolis
    2.00
    Minneapolis
    1.86
     Duluth
    1.64
     MN
    1.47
     Minn
    1.38
     Mn
    1.37
     Fargo
    1.31
     Sioux
    1.31
    Activations Density 0.368%

    No Known Activations