Unverified50% confidenceFactExact time
Existing math reasoning evaluations are almost entirely concentrated on high-resource languages such as English, leaving coverage of low-resource languages—especially Arabic dialects—nearly blank.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/21/2026
First Seen
Valid until: 1/19/2027
Sources
Related Entities
Related Claims
Unverified通用大模型在多层嵌套从句、修辞性倒装等复杂阿拉伯语场景中准确率显著下降75% similarUnverifiedThe scale of annotated corpora for Arabic is far smaller than for English, which constrains the upper bound of purely data-driven approaches.75% similarUnverifiedSafety alignment training data is most comprehensive in English, and may have significant coverage gaps in languages such as Arabic and Urdu, constituting a multilingual jailbreak vulnerability.74% similarUnverifiedفي الإنجليزية يكفي الفصل بالمسافات في الغالب للتجزئة، بينما تعامل النماذج غير المصممة للعربية الكلمة المعقدة كوحدة واحدة مجهولة فتضيع المعلومات الدلالية74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1078822API
curl https://kongchang.com/api/v1/knowledge/claims/1078822MCP
get_claim(id=1078822)