© 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. 24-GEMMASCOPE-2-RES-16K
    7. 1462
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on fragments of demonyms or nationality adjectives (e.g. Bangladeshi, Greek, Bulgarian, Chinese, Azerbaijani, 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_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
     each
    0.54
     Invisalign
    0.54
     A
    0.52
     D
    0.51
     &
    0.51
     taches
    0.51
     L
    0.50
     a
    0.50
     any
    0.48
     Botox
    0.48
    POSITIVE LOGITS
    နိုင်ငံ
    0.57
     парла
    0.54
    geschichte
    0.53
    ⵖ
    0.53
    ชาติ
    0.52
    民主
    0.52
    国的
    0.52
     nación
    0.52
     orientale
    0.52
    ণাস্ত্র
    0.51
    Activations Density 0.005%

    No Known Activations