mlsanityData to 7 Oct 2026

PCL-Reasoner-V1.5: Advancing Math Reasoning with Offline Reinforcement Learning

arXiv 2601.14716 ↗First posted 21 Jan 2026cs.LG cs.AI cs.CLcited by 1

On the map: LLM reasoning → Reinforcement learning for reasoning

also in 1 more
Abstract (arXiv)

We present PCL-Reasoner-V1.5, a 32-billion-parameter large language model (LLM) for mathematical reasoning. The model is built upon Qwen2.5-32B and refined via supervised fine-tuning (SFT) followed by reinforcement learning (RL). A central innovation is our proposed offline RL method, which provides superior training stability and efficiency over standard online RL methods such as GRPO. Our model achieves state-of-the-art performance among models post-trained on Qwen2.5-32B, attaining average accuracies of 90.9% on AIME 2024 and 85.6% on AIME 2025. Our work demonstrates offline RL as a stable and efficient paradigm for advancing reasoning in LLMs. All experiments were conducted on Huawei Ascend 910C NPUs.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Reinforcement learning for reasoning, one of 17 directions in LLM reasoning. Builds on 19 of its 22 references that are in the map. First posted on arXiv 21 Jan 2026.

Builds on

19 of 22 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025A Minimalist Approach to LLM Reasoning: from Rejection Sampling to Reinforce15 Apr
  2. 2025DAPO: An Open-Source LLM Reinforcement Learning System at Scale18 Mar
  3. 2023RAFT: Reward rAnked FineTuning for Generative Foundation Model Alignment13 Apr

Other references · 16

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2023GPT-4 Technical Reportcited by 18,132
  3. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  4. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  5. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  6. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,203
  7. 2020Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problemscited by 1,873
  8. 2025Group Sequence Policy Optimizationcited by 788
  9. 2023Scaling Relationship on Learning Mathematical Reasoning with Large Language Modelscited by 417
  10. 2024Iterative Reasoning Preference Optimizationcited by 244
  11. 2025Skywork Open Reasoner 1 Technical Reportcited by 161
  12. 2025Defeating the Training-Inference Mismatch via FP16cited by 62
4 more
  1. 2025Reinforcement Learning Finetunes Small Subnetworks in Large Language Modelscited by 46
  2. 2025AM-Thinking-v1: Advancing the Frontier of Reasoning at 32B Scalecited by 29
  3. 2025Bridging Offline and Online Reinforcement Learning for LLMscited by 25
  4. 2025When Importance Sampling Misallocates Credit: Asymmetric Ratios for Outcome-Supervised RLcited by 21

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Lightning OPD: Efficient Post-Training for Large Reasoning Models with Offline On-Policy Distillationcited by 28

Similar, not cited

Close in wording, no citation either way

  1. 2025Revisiting Reinforcement Learning for LLM Reasoning from A Cross-Domain Perspective
  2. 2025A Survey of Reinforcement Learning for Large Reasoning Models
  3. 2025Scaling Up RL: Unlocking Diverse Reasoning in LLMs via Prolonged Training
  4. 2025Logic-RL: Unleashing LLM Reasoning with Rule-Based Reinforcement Learning
  5. 2025Reinforcement Learning for Reasoning in Large Language Models with One Training Example
  6. 2025Exploring the Potential of Offline RL for Reasoning in LLMs: A Preliminary Study
  7. 2025AceReason-Nemotron: Advancing Math and Code Reasoning through Reinforcement Learning
  8. 2026Offline Exploration-Aware Fine-Tuning for Long-Chain Mathematical Reasoning