Author name cluster

Lin Yan

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

4 papers

1 author row

NeurIPS Conference 2025 Conference Paper

DAPO: An Open-Source LLM Reinforcement Learning System at Scale

Qiying Yu
Zheng Zhang
Ruofei Zhu
Yufeng Yuan
Xiaochen Zuo
Yu Yue
Weinan Dai
Tiantian Fan

Inference scaling empowers LLMs with unprecedented reasoning ability, with reinforcement learning as the core technique to elicit complex reasoning. However, key technical details of state-of-the-art reasoning LLMs are concealed (such as in OpenAI o1 blog and DeepSeek R1 technical report), thus the community still struggles to reproduce their RL training results. We propose the D ecoupled Clip and D ynamic s A mpling P olicy O ptimization ( DAPO ) algorithm, and fully open-source a state-of-the-art large-scale RL system that achieves 50 points on AIME 2024 using Qwen2. 5-32B base model. Unlike previous works that withhold training details, we introduce four key techniques of our algorithm that make large-scale LLM RL a success. In addition, we open-source our training code, which is built on the verl framework, along with a carefully curated and processed dataset. These components of our open-source system enhance reproducibility and support future research in large-scale LLM RL.

PDF Details

NeurIPS Conference 2025 Conference Paper

EvaLearn: Quantifying the Learning Capability and Efficiency of LLMs via Sequential Problem Solving

Shihan Dou
Ming Zhang
Chenhao Huang
Jiayi Chen
Feng Chen
Shichun Liu
Yan Liu
Chenxiao Liu

We introduce EvaLearn, a pioneering benchmark designed to evaluate large language models (LLMs) on their learning capability and efficiency in challenging tasks, a critical, yet underexplored aspect of model potential. EvaLearn contains 648 challenging problems across six task types, grouped into 182 sequences, each sequence dedicated to one task type. Diverging from most existing benchmarks that evaluate models in parallel, EvaLearn requires models to solve problems sequentially, allowing them to leverage the experience gained from previous solutions. EvaLearn provides five comprehensive automated metrics to evaluate models and quantify their learning capability and efficiency. We extensively benchmark nine frontier models and observe varied performance profiles: some models, such as Claude-3. 7-sonnet, start with moderate initial performance but exhibit strong learning ability, while some models struggle to benefit from experience and may even show negative transfer. Moreover, we investigate model performance under two learning settings and find that instance-level rubrics and teacher-model feedback further facilitate model learning. Importantly, we observe that current LLMs with stronger static abilities do not show a clear advantage in learning capability across all tasks, highlighting that EvaLearn evaluates a new dimension of model performance. We hope EvaLearn provides a novel evaluation perspective for assessing LLM potential and understanding the gap between models and human capabilities, promoting the development of deeper and more dynamic evaluation approaches. All datasets, the automatic evaluation framework, and the results studied in this paper are available in the supplementary materials.

PDF Details

NeurIPS Conference 2025 Conference Paper

Exploring Data Scaling Trends and Effects in Reinforcement Learning from Human Feedback

Wei Shen
Guanlin Liu
Yu Yue
Ruofei Zhu
Qingping Yang
Chao Xin
Lin Yan

Reinforcement Learning from Human Feedback (RLHF) is essential for aligning large language models (LLMs) with human preferences and values. While recent research has primarily focused on algorithmic advancements—such as reducing computational overhead or strengthening reward models to mitigate reward hacking—the critical role of prompt-data construction and its scalability has received comparatively less attention. In this paper, we address this gap by systematically exploring data-driven bottlenecks that currently hinder RLHF performance scaling, focusing specifically on the challenges posed by reward hacking and decreasing response diversity. To mitigate reward hacking, we introduce a hybrid reward system combining reasoning task verifiers (RTV) and a generative reward model (GenRM). This approach not only exhibits enhanced resistance to reward hacking, but also enables accurate assessment of responses against clearly defined ground-truth solutions. Additionally, in order to ensure response diversity and enhance learning effectiveness, we propose a novel prompt-selection method named \textbf{Pre-PPO}, explicitly identifying training prompts that are inherently challenging and thus less prone to reward hacking. Furthermore, we find that \textbf{prioritizing mathematical and coding tasks during the early phases of RLHF training} significantly boosts performance, given that these tasks naturally encode fine-grained response distinctions and possess clearly defined ground truths. Through comprehensive experiments conducted across two model sizes, we validate the effectiveness and scalability of our proposed methods. Results show that RTV exhibits the strongest resistance to reward hacking, followed by GenRM with ground truth, and finally GenRM relying on SFT Best-of-N responses. Moreover, our proposed strategies enable the model to rapidly capture subtle task-specific distinctions, leading to substantial improvements in overall RLHF performance. This work underscores the importance of careful data construction and provides practical methodologies to overcome critical performance barriers in RLHF.

PDF Details

EAAI Journal 2025 Journal Article

Multi-attribute decision-making with extended VlseKriterijuska Optimizacija I Komoromisno Resenje method based on generalized Fermatean probabilistic hesitant fuzzy distance measure and cross-entropy

Chuanyang Ruan
Lin Yan
Shouzhen Zeng

Amidst the escalating complexity and ambiguity inherent in real-world decision-making within uncertain environments, the Fermatean probabilistic hesitant fuzzy set (FPHFS) emerges as a significant extension of the Fermatean fuzzy set (FFS), gaining momentum in the realm of multi-attribute decision-making (MADM). This paper proposes a novel MADM approach that integrates the distance measure and cross-entropy of FPHFSs with the VlseKriterijuska Optimizacija I Komoromisno Resenje (VIKOR) method. This approach is designed to concurrently account for the objective weighting of attributes, the stochastic nature of evaluation data, and the psychological preferences of decision-maker (DM). Initially, leveraging the definitions of distance measures and entropy for FPHFSs, various forms of distance and cross-entropy measures are proposed. Subsequently, by integrating the proposed generalized Fermatean probabilistic hesitant fuzzy cross-entropy with the maximum deviation method, two models for determining the objective weights of attributes are developed. To intuitively capture and process DMs' preference information, a linear preference function within the decision-making process is constructed using the introduced generalized Fermatean probabilistic hesitant fuzzy distance. Ultimately, the preference function-based Fermatean probabilistic hesitant fuzzy VIKOR (FPHF-VIKOR) approach is established. By applying it to a numerical example of supply chain finance risk assessment, the decision-making procedure of the method is thoroughly demonstrated. The excellent performance of the distance and cross-entropy measures within this model is examined through parameter sensitivity analysis. Additionally, a real-world application concerning financing enterprise selection highlights the practicality and scalability of the proposed method. Comparative analysis with existing MADM methods and VIKOR variants further confirms the scientific feasibility and effectiveness of our approach.

Details DOI