跳到正文
arXiv cs.AI· Nicol\'as Astorga, Anita Kriz, Mihaela van der Schaar·· 4 小时前AI 评分38

CauSim:用日益复杂的因果模拟器扩展 LLM 因果推理能力

Scaling Causal Reasoning with Increasingly Complex Causal Simulators

AI 导读

研究者提出 CauSim 框架,将因果推理从标签稀缺问题转化为可扩展的监督学习问题。该框架由 LLM 增量构建可执行的结构因果模型(SCM),在扩展到全局复杂系统的同时保持因果查询答案可验证,并能将可执行 SCM 翻译为自然语言以实现跨表示监督。实验显示其具备跨表示泛化能力,课程规模与数据量带来一致增益,并支持 LLM 通过自生成模拟器自我改进。

正文 · 原文

View PDF HTML (experimental)

Abstract:Despite surpassing human performance across mathematics, coding, and other knowledge-intensive tasks, large language models (LLMs) continue to struggle with causal reasoning. A core obstacle is the target data itself: causal systems are complex and often expressed in non-executable forms, while ground-truth answers to causal queries are inherently scarce. We introduce CauSim, a framework that turns causal reasoning from a scarce-label problem into a scalable supervised one. CauSim constructs increasingly complex causal simulators: executable structural causal models (SCMs), incrementally built by LLMs, that scale to globally complex systems while maintaining verifiable answers to causal queries. CauSim operates across representations by formalizing non-executable causal knowledge into code, enabling data augmentation, and translating executable SCMs into natural language, enabling supervision in previously difficult-to-supervise representations. We structure our research into two parts: (1) how to construct increasingly complex causal simulators, and (2) a systematic study of what CauSim enables, demonstrating generalization across representations, consistent gains from curriculum scaling and data volume, LLM self-improvement through self-generated simulators, and data augmentation via formalization of existing domain knowledge.
Subjects: Artificial Intelligence (cs.AI)
Cite as: arXiv:2605.09079 [cs.AI]
  (or arXiv:2605.09079v2 [cs.AI] for this version)
  https://doi.org/10.48550/arXiv.2605.09079

arXiv-issued DOI via DataCite

Submission history

From: Nicolas Astorga [view email]
[v1] Sat, 9 May 2026 17:39:26 UTC (661 KB)
[v2] Tue, 17 Nov 2026 00:54:34 UTC (2,200 KB)

来源:arXiv cs.AI · arxiv.org