During neural machine translation, why might a model using beam search produce grammatically incorrect output despite high BLEU scores on the training data?

💡 Explanation

The model might prioritize short, precise n-grams reflected in the BLEU score, rather than capturing long-range dependencies needed for grammatical correctness, because the BLEU score primarily assesses n-gram precision. Therefore, the model produces grammatically incorrect output, rather than focusing on semantic correctness or fluency, when maximizing BLEU score.

• A museum visitor points toward a specific painting while describing its historical context to a friend. Which mechanism is predominantly at play in their nonverbal communication?
• In Tokyo Japanese, if a speaker utters the phrase "hashi" (bridge) immediately before "ga" (subject marker) during casual conversation, which consequence follows?
• Why does intergenerational language transmission fail in immigrant communities despite parental proficiency?
• Why does multimodal sentiment analysis of customer service interactions often misinterpret sarcasm?
• A politician uses 'doublespeak' during a press conference to obscure the details of a controversial policy. Which mechanism enables this persuasive manipulation?
• What causes increased reading difficulty when encountering a logographic writing system, like ancient Chinese characters, after mastering an alphabetic system?