CALM Before the STORM: Unlocking Native Reasoning for Optimization Modeling

ICLR 2026 Conference Submission25082 Authors

Keywords: Large Reasoning Models, Tool Use, Domain Adaptation, Reasoning Alignment, Optimization Modeling

TL;DR: How do you get a 4B model to perform like a 671B giant? Don't force it, guide it. Our CALM framework uses gentle hints to teach a small LRM to think smart, before unleashing its full potential with RL to create STORM.

Abstract: Large Reasoning Models (LRMs) have demonstrated strong capabilities in complex multi-step reasoning, opening new opportunities for automating optimization modeling. However, existing domain adaptation methods, originally designed for earlier instruction-tuned models, often fail to exploit the advanced reasoning patterns of modern LRMs --- In particular, we show that direct fine-tuning on traditional non-reflective datasets leads to limited gains. To fully leverage LRMs’ inherent reasoning abilities, we propose **CALM** (Corrective Adaptation with Lightweight Modification), a framework that progressively refines LRMs within their native reasoning modes for optimization modeling tasks. In CALM, an expert intervener identifies reasoning flaws and provides concise corrective hints, which the LRM incorporates to produce improved reasoning trajectories. These interventions modify fewer than 2.6% of generated tokens, but generate high-quality data for soft adaptation through supervised fine-tuning. The adapted model is then further improved through reinforcement learning. Building on CALM, we develop **STORM** (Smart Thinking Optimization Reasoning Model), a 4B-parameter LRM that achieves a new state-of-the-art average accuracy of 68.9% across five popular optimization modeling benchmarks, matching the performance of a 671B LRM. These results demonstrate that dynamic, hint-based data synthesis both preserves and amplifies the native reasoning patterns of modern LRMs, offering a more effective and scalable path towards expert-level performance on challenging optimization modeling tasks.

Primary Area: foundation or frontier models, including LLMs

Submission Number: 25082