Aligning with Logic: Measuring, Evaluating and Improving Logical Preference Consistency in Large Language Models
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913684329070592 |
|---|---|
| author | Liu, Yinhong Guo, Zhijiang Liang, Tianya Shareghi, Ehsan Vulić, Ivan Collier, Nigel |
| author_facet | Liu, Yinhong Guo, Zhijiang Liang, Tianya Shareghi, Ehsan Vulić, Ivan Collier, Nigel |
| contents | Large Language Models (LLMs) are expected to be predictable and trustworthy to support reliable decision-making systems. Yet current LLMs often show inconsistencies in their judgments. In this work, we examine logical preference consistency as a foundational requirement for building more dependable LLM systems, ensuring stable and coherent decision-making while minimizing erratic or contradictory outputs. To quantify the logical preference consistency, we propose a universal evaluation framework based on three fundamental properties: transitivity, commutativity and negation invariance. Through extensive experimentation across diverse LLMs, we demonstrate that these properties serve as strong indicators of judgment robustness. Furthermore, we introduce a data refinement and augmentation technique, REPAIR, that enhances logical consistency while maintaining alignment with human preferences. Finally, we show that improving consistency leads to better performance in LLM-driven logic-based algorithms, reinforcing stability and coherence in decision-making systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_02205 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Aligning with Logic: Measuring, Evaluating and Improving Logical Preference Consistency in Large Language Models Liu, Yinhong Guo, Zhijiang Liang, Tianya Shareghi, Ehsan Vulić, Ivan Collier, Nigel Computation and Language Artificial Intelligence Logic in Computer Science Large Language Models (LLMs) are expected to be predictable and trustworthy to support reliable decision-making systems. Yet current LLMs often show inconsistencies in their judgments. In this work, we examine logical preference consistency as a foundational requirement for building more dependable LLM systems, ensuring stable and coherent decision-making while minimizing erratic or contradictory outputs. To quantify the logical preference consistency, we propose a universal evaluation framework based on three fundamental properties: transitivity, commutativity and negation invariance. Through extensive experimentation across diverse LLMs, we demonstrate that these properties serve as strong indicators of judgment robustness. Furthermore, we introduce a data refinement and augmentation technique, REPAIR, that enhances logical consistency while maintaining alignment with human preferences. Finally, we show that improving consistency leads to better performance in LLM-driven logic-based algorithms, reinforcing stability and coherence in decision-making systems. |
| title | Aligning with Logic: Measuring, Evaluating and Improving Logical Preference Consistency in Large Language Models |
| topic | Computation and Language Artificial Intelligence Logic in Computer Science |
| url | https://arxiv.org/abs/2410.02205 |