Saved in:
Bibliographic Details
Main Authors: Kumar, Medha, Xu, Zifei, Wang, Xin, Webb, Tristan
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.15694
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • Strong reasoning capabilities can now be achieved by large-scale reinforcement learning (RL) without any supervised fine-tuning. Although post-training quantization (PTQ) and quantization-aware training (QAT) are well studied in the context of fine-tuning, how quantization impacts RL in large reasoning models (LRMs) remains an open question. To answer this question, we conducted systematic experiments and discovered a significant gap in reasoning performance on mathematical benchmarks between post-RL quantized models and their quantization-aware RL optimized counterparts. Our findings suggest that quantization-aware RL training negatively impacted the learning process, whereas PTQ and QLoRA led to greater performance.