Unverified70% confidenceFactExact time
Domestic models such as GLM have increased the proportion of high-quality Chinese corpus during the training phase and optimized the tokenizer design for Chinese characters.
2
Sources
70%
Confidence
Long-term
Relevance
9/3/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedGLM adopys an autoregressive blank infilling pretraining method and demonstrates outstanding performance in bilingual Chinese-English understanding and generation.74% similarVerifiedLLMs acquire language capabilities by performing unsupervised pre-training on massive text corpora (predicting the next token).70% similarUnverified粤语书写系统至今未完全标准化，同一口语词汇存在多种汉字写法，导致模型训练面临标签噪声问题70% similarUnverified豆包输入法背后依托字节跳动的语音AI技术积累，针对中文技术术语和混合中英文场景做了专项优化68% similarUnverified主流扩散模型的文字生成依赖CLIP等视觉-语言预训练模型，其训练数据以英文为主，导致中文字符生成困难67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/849941API
curl https://kongchang.com/api/v1/knowledge/claims/849941MCP
get_claim(id=849941)