${\pi }^2$: Structure-Originated Reasoning Data Improves Long-Context Reasoning Ability of Large Language Models

We study a pipeline that curates reasoning data from initial structured data for improving long-context reasoning in large language models (LLMs). Our approach, ${\pi }^2$, constructs high-quality reasoning data through rigorous QA curation: 1) extracting and expanding tables from Wikipedia, 2) from the collected tables and relevant context, generating realistic and multi-hop analytical reasoning questions whose answers are automatically determined and verified through dual-path code execution, and 3) back-translating step-by-step structured reasoning traces as solutions of QA pairs given realistic web-search context. Supervised fine-tuning with $\text{textsc}gpt-oss-20b$ and $\text{textsc}Qwen3-4B-Instruct-2507$ on ${\pi }^2$ yields consistent improvements across four long-context reasoning benchmarks and our alike ${\pi }^2$-Bench, with average absolute accuracy gains of +4.3% and +2.7% respectively. Notably, our dataset facilitates self-distillation, where $\text{textsc}gpt-oss-20b$ even improves its average performance by +4.4% with its own reasoning traces, demonstrating ${\pi }^2$’s usefulness. Our code, data, and models are open-source at https://github.com/vt-pi-squared/pi-squared.