QwenLong-L1: Towards Long-Context Large Reasoning Models with Reinforcement Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wan, Fanqi, Shen, Weizhou, Liao, Shengyi, Shi, Yingcheng, Li, Chenliang, Yang, Ziyi, Zhang, Ji, Huang, Fei, Zhou, Jingren, Yan, Ming
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912397320519680
author Wan, Fanqi
Shen, Weizhou
Liao, Shengyi
Shi, Yingcheng
Li, Chenliang
Yang, Ziyi
Zhang, Ji
Huang, Fei
Zhou, Jingren
Yan, Ming
author_facet Wan, Fanqi
Shen, Weizhou
Liao, Shengyi
Shi, Yingcheng
Li, Chenliang
Yang, Ziyi
Zhang, Ji
Huang, Fei
Zhou, Jingren
Yan, Ming
contents Recent large reasoning models (LRMs) have demonstrated strong reasoning capabilities through reinforcement learning (RL). These improvements have primarily been observed within the short-context reasoning tasks. In contrast, extending LRMs to effectively process and reason on long-context inputs via RL remains a critical unsolved challenge. To bridge this gap, we first formalize the paradigm of long-context reasoning RL, and identify key challenges in suboptimal training efficiency and unstable optimization process. To address these issues, we propose QwenLong-L1, a framework that adapts short-context LRMs to long-context scenarios via progressive context scaling. Specifically, we utilize a warm-up supervised fine-tuning (SFT) stage to establish a robust initial policy, followed by a curriculum-guided phased RL technique to stabilize the policy evolution, and enhanced with a difficulty-aware retrospective sampling strategy to incentivize the policy exploration. Experiments on seven long-context document question-answering benchmarks demonstrate that QwenLong-L1-32B outperforms flagship LRMs like OpenAI-o3-mini and Qwen3-235B-A22B, achieving performance on par with Claude-3.7-Sonnet-Thinking, demonstrating leading performance among state-of-the-art LRMs. This work advances the development of practical long-context LRMs capable of robust reasoning across information-intensive environments.
format Preprint
id arxiv_https___arxiv_org_abs_2505_17667
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle QwenLong-L1: Towards Long-Context Large Reasoning Models with Reinforcement Learning
Wan, Fanqi
Shen, Weizhou
Liao, Shengyi
Shi, Yingcheng
Li, Chenliang
Yang, Ziyi
Zhang, Ji
Huang, Fei
Zhou, Jingren
Yan, Ming
Computation and Language
Recent large reasoning models (LRMs) have demonstrated strong reasoning capabilities through reinforcement learning (RL). These improvements have primarily been observed within the short-context reasoning tasks. In contrast, extending LRMs to effectively process and reason on long-context inputs via RL remains a critical unsolved challenge. To bridge this gap, we first formalize the paradigm of long-context reasoning RL, and identify key challenges in suboptimal training efficiency and unstable optimization process. To address these issues, we propose QwenLong-L1, a framework that adapts short-context LRMs to long-context scenarios via progressive context scaling. Specifically, we utilize a warm-up supervised fine-tuning (SFT) stage to establish a robust initial policy, followed by a curriculum-guided phased RL technique to stabilize the policy evolution, and enhanced with a difficulty-aware retrospective sampling strategy to incentivize the policy exploration. Experiments on seven long-context document question-answering benchmarks demonstrate that QwenLong-L1-32B outperforms flagship LRMs like OpenAI-o3-mini and Qwen3-235B-A22B, achieving performance on par with Claude-3.7-Sonnet-Thinking, demonstrating leading performance among state-of-the-art LRMs. This work advances the development of practical long-context LRMs capable of robust reasoning across information-intensive environments.
title QwenLong-L1: Towards Long-Context Large Reasoning Models with Reinforcement Learning
topic Computation and Language
url https://arxiv.org/abs/2505.17667