mlsanityData to 7 Oct 2026

Externalized CPDAG Summaries Improve LLM Causal Deduction

arXiv 2609.31071 ↗First posted 25 Sep 2026cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Corr2Cause asks whether a causal claim holds in every DAG compatible with observed correlations and conditional independencies. We frame this as latent-object reasoning: the label is defined by a CPDAG query, but free-form chain-of-thought often collapses the Markov-equivalence-class problem into local pattern matching. We propose Structured Thinking, a two-turn pipeline that first externalizes a typed, schema-constrained CPDAG summary and then answers against that graph state. On the Corr2Cause full test, Structured Thinking raises Qwen3.5-27B from 73.0 to 86.4 F1(Yes) over a strong PC-instruction baseline in the primary paired run (+13.4 pp; McNemar p=2.4× 10-6; bootstrap 95% CI [+8.4, +18.6]); across three full-ID seeds, the mean gain is +8.1 ± 5.3 pp. A PC-scaffolded two-turn prose control reaches only 67.6 F1, indicating that a detailed PC scaffold plus a schema-free prose intermediate is not sufficient. The same pattern holds on Qwen3.6-27B, Paraphrase-OOD, and GPT-5.4-mini. Scrambling the emitted CPDAG costs 12.0 pp F1, and a full-split audit shows close agreement with the reference CPDAG (ID skeleton F1 0.960; exact match 75.9%). These results support a bounded design principle: externalize the latent object that defines the label, constrain its form, and test whether downstream answers use it.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 23 of its 30 references that are in the map. First posted on arXiv 25 Sep 2026.

Builds on

23 of 30 references.

References · 23

  1. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  2. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
  3. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,072
  4. 2022Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Taskscited by 1,057
  5. 2022PAL: Program-aided Language Modelscited by 635
  6. 2023Causal Reasoning and Large Language Models: Opening a New Frontier for Causalitycited by 348
  7. 2022TALM: Tool Augmented Language Modelscited by 197
  8. 2023Efficient Guided Generation for Large Language Modelscited by 192
  9. 2013Causal Inference and Causal Explanation with Background Knowledgecited by 168
  10. 2023Can Large Language Models Infer Causation from Correlation?cited by 142
  11. 2023Can large language models build causal graphs?cited by 72
  12. 2025Unveiling Causal Reasoning in Large Language Models: Reality or Mirage?cited by 36
11 more
  1. 2024Causal Inference with Large Language Model: A Surveycited by 31
  2. 2022Can Foundation Models Talk Causality?cited by 30
  3. 2025Causal Sufficiency and Necessity Improves Chain-of-Thought Reasoningcited by 16
  4. 2024A Critical Review of Causal Reasoning Benchmarks for Large Language Modelscited by 13
  5. 2024C2P: Featuring Large Language Models with Causal Reasoningcited by 8
  6. 2024Prompting Strategies for Enabling Large Language Models to Infer Causation from Correlationcited by 7
  7. 2026C2-Faith: Benchmarking LLM Judges for Causal and Coverage Faithfulness in Chain-of-Thought Reasoningcited by 6
  8. 2025Causal Reasoning in Pieces: Modular In-Context Learning for Causal Discoverycited by 3
  9. 2025Structured Thinking Matters: Improving LLMs Generalization in Causal Inference Taskscited by 2
  10. 2026Bypassing the Rationale: Causal Auditing of Implicit Reasoning in Language Modelscited by 2
  11. 2025CLEAR-3K: Assessing Causal Explanatory Capabilities in Language Modelscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025KisMATH: Do LLMs Have Knowledge of Implicit Structures in Mathematical Reasoning?
  2. 2024CausalGraph2LLM: Evaluating LLMs for Causal Queries
  3. 2026Thinking Leakage: A Causal Audit of NoThink Post-Training in Hybrid Reasoning Models
  4. 2025Unveiling and Causalizing CoT: A Causal Pespective
  5. 2026CauSim: Scaling Causal Reasoning with Increasingly Complex Causal Simulators
  6. 2025Compositional Causal Reasoning Evaluation in Language Models
  7. 2026Breaking the Chain: A Causal Analysis of LLM Faithfulness to Intermediate Structures
  8. 2026From Chains to DAGs: Probing the Graph Structure of Reasoning in LLMs