Unverified50% confidenceFactExact time
Traditional n-gram models and neural language models trained on a single language assign extremely low probability scores to cross-lingual word sequences, causing the decoder to tend to forcibly map foreign-language words to phonetically similar words in the target language.
1
Sources
50%
Confidence
Long-term
Relevance
8/8/2026
First Seen
Sources
语码转换ASR评测：WER为何失灵及替代方案
Related Claims
Unverified多语言预训练模型如mBERT、XLM-R展现出跨语言零样本迁移能力，即便某些语言训练中从未与其他语言平行语料共同出现78% similarUnverifiedThe language models used in traditional speech recognition are usually lightweight N-gram models or small RNNs with limited capability.77% similarUnverifiedPerplexity measures a language model's ability to predict test text; the lower the value, the less surprised the model is by the text.76% similarUnverifiedModern multilingual large models establish language-agnostic semantic representations during the pre-training phase, building cross-lingual semantic alignment in high-dimensional vector space through joint training on massive multilingual corpora.76% similarUnverifiedOOD detection for autoregressive language models is more difficult than for traditional machine learning because the linguistic compositional space is vast, semantic OOD does not correspond to surface-form OOD, and adversaries can construct inputs that appear statistically in-distribution but semantically trigger OOD behavior.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/707043API
curl https://kongchang.com/api/v1/knowledge/claims/707043MCP
get_claim(id=707043)