Reasoning Pattern Matters: Learning to Reason without Human Rationales
Chaoxu Pang, Yixuan Cao, Ping Luo
TL;DR
This work shows that in patterned reasoning tasks, the key to effective reasoning supervision is learning a fixed reasoning pattern rather than collecting large volumes of high quality rationales. It introduces PARO, a framework that lets large language models automatically generate rationale traces aligned with task specific patterns, dramatically reducing human annotation costs. Through two representative tasks, numerical semantic matching and transaction purpose classification, PARO achieves comparable or superior performance to human annotated baselines while using an order of magnitude fewer rationales. The findings suggest a practical path to scalable reasoning supervision by emphasizing pattern enforcement over data quantity or rationale correctness, with implications for RLVR pipelines and real-world decision workflows.
Abstract
Large Language Models (LLMs) have demonstrated remarkable reasoning capabilities under the widely adopted SFT+RLVR paradigm, which first performs Supervised Fine-Tuning (SFT) on human-annotated reasoning trajectories (rationales) to establish initial reasoning behaviors, then applies Reinforcement Learning with Verifiable Rewards (RLVR) to optimize the model using verifiable signals without golden rationales. However, annotating high-quality rationales for the SFT stage remains prohibitively expensive. This paper investigates when and how rationale annotation costs can be substantially reduced without compromising reasoning performance. We identify a broad class of problems, termed patterned reasoning tasks, where reasoning follows a fixed, procedural strategy consistent across instances. Although instances vary in content such as domain knowledge, factual information, or numeric values, the solution derives from applying a shared reasoning pattern. We argue that the success of SFT+RLVR on such tasks primarily stems from its ability to enable models to internalize these reasoning patterns. Using numerical semantic matching as a representative task, we provide both causal and behavioral evidence showing that reasoning patterns rather than the quantity or quality of rationales are the key determinant of performance. Building on these insights, we propose Pattern-Aware LLMs as Rationale AnnOtators (PARO), a simple yet effective framework that enables LLMs to generate rationales aligned with task-specific reasoning patterns without requiring human rationale annotations. Experiments show that PARO-generated rationales achieve comparable SFT+RLVR performance to human rationales that are 10 times larger. These results suggest that large-scale human rationale annotations can be replaced with LLM-based automatic annotations requiring only limited human supervision over reasoning patterns.