mlsanityData to 14 Oct 2026

Investigating the (De)Composition Capabilities of Large Language Models in Natural-to-Formal Language Conversion

arXiv 2501.14649 ↗First posted 24 Jan 2025cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

To achieve generalized and robust natural-to-formal language conversion (N2F), large language models (LLMs) need to have strong capabilities of decomposition and composition in N2F when faced with an unfamiliar formal language and be able to cope with compositional gaps and counter-intuitive symbolic names. To investigate whether LLMs have this set of basic capabilities in N2F, we propose the DEDC framework. This framework semi-automatically performs sample and task construction, allowing decoupled evaluation of the set of decomposition and composition capabilities of LLMs in N2F. Based on this framework, we evaluate and analyze the most advanced LLMs, and the main findings include that: (1) the LLMs are deficient in both decomposition and composition; (2) the LLMs show a wide coverage of error types that can be attributed to deficiencies in natural language understanding and the learning and use of symbolic systems; (3) compositional gaps and counter-intuitive symbolic names both affect the decomposition and composition of the LLMs. Our work provides a new perspective for investigating the basic capabilities of decomposition and composition of LLMs in N2F. The detailed analysis of deficiencies and attributions can help subsequent improvements of LLMs.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 21 of its 35 references that are in the map. First posted on arXiv 24 Jan 2025.

Builds on

21 of 35 references.

References · 21

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2022A Survey on In-context Learningcited by 721
  3. 2024A Survey on Large Language Models for Code Generationcited by 658
  4. 2019TabFact: A Large-scale Dataset for Table-based Fact Verificationcited by 428
  5. 2016Deep Biaffine Attention for Neural Dependency Parsingcited by 334
  6. 2019Measuring Compositional Generalization: A Comprehensive Method on Realistic Datacited by 284
  7. 2019Compositionality decomposed: how do neural networks generalise?cited by 265
  8. 2020COGS: A Compositional Generalization Challenge Based on Semantic Interpretationcited by 263
  9. 2023Large Language Models are Versatile Decomposers: Decompose Evidence and Questions for Table-based Reasoningcited by 178
  10. 2020Compositional Generalization and Natural Language Variation: Can a Semantic Parsing Approach Handle Both?cited by 147
  11. 2022Diverse Demonstrations Improve In-context Compositional Generalizationcited by 139
  12. 2024Faithful Logical Reasoning via Symbolic Chain-of-Thoughtcited by 100
9 more
  1. 2024SciAgent: Tool-augmented Language Models for Scientific Reasoningcited by 70
  2. 2024A Survey on Employing Large Language Models for Text-to-SQL Taskscited by 67
  3. 2023How Do In-Context Examples Affect Compositional Generalization?cited by 51
  4. 2020Logic2Text: High-Fidelity Natural Language Generation from Logical Formscited by 45
  5. 2020A Survey of Syntactic-Semantic Parsing Based on Constituent and Dependency Structurescited by 7
  6. 2024How Proficient Are Large Language Models in Formal Languages? An In-Depth Insight for Knowledge Base Question Answeringcited by 5
  7. 2024SPOR: A Comprehensive and Practical Evaluation Method for Compositional Generalization in Data-to-Text Generationcited by 3
  8. 2024Structured Tree Alignment for Evaluation of (Speech) Constituency Parsingcited by 1
  9. 2024LLMs' morphological analyses of complex FST-generated Finnish wordscited by 1

Built on by

Not cited yet.