© 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. 8974
    Prev
    Next
    INDEX
    Explanations

    an

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

    The neuron activates on place names and geographic location tokens (e.g. “US Virgin Islands,” “St. John,” etc.).

    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
    県
    0.60
     Lincolnshire
    0.60
     tỉnh
    0.59
     англий
    0.58
    Queensland
    0.57
     ইংল্যান্ড
    0.57
     angiel
    0.57
    イング
    0.56
     県
    0.55
     maaf
    0.55
    POSITIVE LOGITS
     Federal
    0.93
     congressional
    0.92
     federal
    0.92
     Congressional
    0.92
    Congressional
    0.91
     Puerto
    0.85
    Federal
    0.84
    federal
    0.84
     федера
    0.83
     federally
    0.80
    Activations Density 0.227%

    No Known Activations