Unverified50% confidenceFactExact time
The study analyzed the word lists of twelve alphabetic writing systems and reported results in detail for five of these languages.
1
Sources
50%
Confidence
Long-term
Relevance
11/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedA study based on a corpus containing 131,000 sentences covering 11 Indian scripts compared the tokenizer performance of models such as TinyLlama-1.1B, mBERT, and MuRIL.79% similarUnverifiedThe study used eye-tracking data from participants with 12 different native language (L1) backgrounds, with the reading texts uniformly in English.73% similarUnverifiedThe study tested six mainstream large language models.72% similarUnverifiedThe study tested 12 models using psycholinguistic datasets and generic statements from natural corpora.71% similarUnverifiedThe researchers used the MTEB and MMTEB benchmarks, covering 18 languages, and conducted more than 200 monolingual evaluations.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1158286API
curl https://kongchang.com/api/v1/knowledge/claims/1158286MCP
get_claim(id=1158286)