What Makes Low-Bit Quantization-Aware Training Work for Reasoning LLMs? A Systematic Study
This addresses the efficiency challenge of slow inference in reasoning LLMs for applications like coding and mathematics, but it is incremental as it builds on existing quantization techniques.
The study tackled the problem of low-bit quantization for reasoning large language models (LLMs), which often suffer from accuracy drops, and found that an optimized quantization-aware training workflow consistently outperforms state-of-the-art post-training quantization methods, achieving gains such as 44.53% on MATH-500 for a specific model.
Reasoning models excel at complex tasks such as coding and mathematics, yet their inference is often slow and token-inefficient. To improve the inference efficiency, post-training quantization (PTQ) usually comes with the cost of large accuracy drops, especially for reasoning tasks under low-bit settings. In this study, we present a systematic empirical study of quantization-aware training (QAT) for reasoning models. Our key findings include: (1) Knowledge distillation is a robust objective for reasoning models trained via either supervised fine-tuning or reinforcement learning; (2) PTQ provides a strong initialization for QAT, improving accuracy while reducing training cost; (3) Reinforcement learning remains feasible for quantized models given a viable cold start and yields additional gains; and (4) Aligning the PTQ calibration domain with the QAT training domain accelerates convergence and often improves the final accuracy. Finally, we consolidate these findings into an optimized workflow (Reasoning-QAT), and show that it consistently outperforms state-of-the-art PTQ methods across multiple LLM backbones and reasoning datasets. For instance, on Qwen3-0.6B, it surpasses GPTQ by 44.53% on MATH-500 and consistently recovers performance in the 2-bit regime.