A Survey of LLM Alignment: Instruction Understanding, Intention Reasoning, and Reliable Generation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912858562887680 |
|---|---|
| author | Chang, Zongyu Lu, Feihong Zhu, Ziqin Li, Qian Ji, Cheng Yang, Tao Chen, Zhuo Peng, Hao Liu, Yang Xu, Ruifeng Song, Yangqiu Li, Jianxin Wang, Shangguang |
| author_facet | Chang, Zongyu Lu, Feihong Zhu, Ziqin Li, Qian Ji, Cheng Yang, Tao Chen, Zhuo Peng, Hao Liu, Yang Xu, Ruifeng Song, Yangqiu Li, Jianxin Wang, Shangguang |
| contents | Large language models have demonstrated exceptional capabilities in understanding and generation. However, in real-world scenarios, users' natural language expressions are often inherently fuzzy, ambiguous, and uncertain, leading to challenges such as vagueness, polysemy, and contextual ambiguity. This paper focuses on three challenges in LLM-based text generation tasks: instruction understanding, intention reasoning, and reliable dialog generation. Regarding human complex instruction, LLMs have deficiencies in understanding long contexts and instructions in multi-round conversations. For intention reasoning, LLMs may have inconsistent command reasoning, difficulty reasoning about commands containing incorrect information, difficulty understanding user ambiguous language commands, and a weak understanding of user intention in commands. Besides, In terms of Reliable Dialog Generation, LLMs may have unstable generated content and unethical generation. To this end, we classify and analyze the performance of LLMs in challenging scenarios and conduct a comprehensive evaluation of existing solutions. Furthermore, we introduce benchmarks and categorize them based on the aforementioned three core challenges. Finally, we explore potential directions for future research to enhance the reliability and adaptability of LLMs in real-world applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_09101 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | A Survey of LLM Alignment: Instruction Understanding, Intention Reasoning, and Reliable Generation Chang, Zongyu Lu, Feihong Zhu, Ziqin Li, Qian Ji, Cheng Yang, Tao Chen, Zhuo Peng, Hao Liu, Yang Xu, Ruifeng Song, Yangqiu Li, Jianxin Wang, Shangguang Human-Computer Interaction Large language models have demonstrated exceptional capabilities in understanding and generation. However, in real-world scenarios, users' natural language expressions are often inherently fuzzy, ambiguous, and uncertain, leading to challenges such as vagueness, polysemy, and contextual ambiguity. This paper focuses on three challenges in LLM-based text generation tasks: instruction understanding, intention reasoning, and reliable dialog generation. Regarding human complex instruction, LLMs have deficiencies in understanding long contexts and instructions in multi-round conversations. For intention reasoning, LLMs may have inconsistent command reasoning, difficulty reasoning about commands containing incorrect information, difficulty understanding user ambiguous language commands, and a weak understanding of user intention in commands. Besides, In terms of Reliable Dialog Generation, LLMs may have unstable generated content and unethical generation. To this end, we classify and analyze the performance of LLMs in challenging scenarios and conduct a comprehensive evaluation of existing solutions. Furthermore, we introduce benchmarks and categorize them based on the aforementioned three core challenges. Finally, we explore potential directions for future research to enhance the reliability and adaptability of LLMs in real-world applications. |
| title | A Survey of LLM Alignment: Instruction Understanding, Intention Reasoning, and Reliable Generation |
| topic | Human-Computer Interaction |
| url | https://arxiv.org/abs/2502.09101 |