🔭 Research Overview

My research aims at building agentic methods that operate real-world software, along two threads:

🧵 Publication Threads

^* equal contribution  ·  ^† project co-lead

• Preprint
  OSReward: Instituting Standardized Evaluation for Cross-Platform Computer-Use Reward Models

  Qiushi Sun^†, Kanzhi Cheng^†, Yian Wang, Bowen Yang, Hang Yan, Liheng Chen, Fangzhi Xu, Zichen Ding, Nuo Chen, Jialin Cao, Xingdong Gong, Zehao Li, Kaiming Jin, Xinfeng Yuan, Zhoumianze Liu, Jingyang Gong, Zhangyue Yin, Jiahui Gao, Zhiyong Wu, Tianbao Xie, Jianbing Zhang, Ben Kao, Lingpeng Kong

  Paper Project OSReward OS-Shepherd-9B/35B Data Code BIB

  A standardized benchmark for computer-use reward models, with OS-Shepherd judges trained on 100K trajectory judgments.
• ICLR'26
  ScienceBoard: Evaluating Multimodal Autonomous Agents in Realistic Scientific Workflows

  Qiushi Sun, Zhoumianze Liu, Chang Ma, Zichen Ding, Fangzhi Xu, Zhangyue Yin, Haiteng Zhao, Zhenyu Wu, Kanzhi Cheng, Zhaoyang Liu, Jianing Wang, Qintong Li, Xiangru Tang, Tianbao Xie, Xiachong Feng, Xiang Li, Ben Kao, Wenhai Wang, Biqing Qi, Lingpeng Kong, Zhiyong Wu

  Paper Slide Project HF Env Code BIB

  A realistic environment and benchmark for evaluating computer-using agents in scientific workflows.
• ICLR'26
  JanusCoder: Towards a Foundational Visual-Programmatic Interface for Code Intelligence

  Qiushi Sun^*, Jingyang Gong^*, Yang Liu^*, Qiaosheng Chen^*, Lei Li, Kai Chen, Qipeng Guo, Ben Kao, Fei Yuan

  Paper Slide Project HF Data Code BIB

  Foundation models that establish a unified visual-programmatic interface for code intelligence.
• ACL'26 Oral
  OS-Sentinel: Towards Safety-Enhanced Mobile GUI Agents via Hybrid Validation in Realistic Workflows

  Qiushi Sun^*, Mukai Li^*, Zhoumianze Liu^*, Zhihui Xie^*, Fangzhi Xu, Zhangyue Yin, Kanzhi Cheng, Zehao Li, Zichen Ding, Qi Liu, Zhiyong Wu, Zhuosheng Zhang, Ben Kao, Lingpeng Kong

  Paper Slide Project Code BIB

  🏆 AIWILD @ ICLR 2026 Best Paper Award

  A hybrid safety detection framework for mobile GUI agents, pairing formal verification with contextual judgment.
• ACL'26 Oral
  CodeEvo: Interaction-Driven Synthesis of Code-centric Data through Hybrid and Iterative Feedback

  Qiushi Sun, Jingyang Gong, Lei Li, Qipeng Guo, Fei Yuan

  Paper Slide Code Data BIB

  Synthesizes code-centric training data through coder–reviewer agent interaction with hybrid feedback.
• ACL'25
  OS-Genesis: Automating GUI Agent Trajectory Construction via Reverse Task Synthesis

  Qiushi Sun^*, Kanzhi Cheng^*, Zichen Ding^*, Chuanyang Jin^*, Yian Wang, Fangzhi Xu, Zhenyu Wu, Chengyou Jia, Liheng Chen, Zhoumianze Liu, Ben Kao, Guohao Li, Junxian He, Yu Qiao, Zhiyong Wu

  Paper Slide Project HF Data Code BIB

  Constructs high-quality GUI agent trajectories without human supervision via reverse task synthesis.
• COLM'26
  OpenMobile: Building Open Mobile Agents with Task and Trajectory Synthesis

  Kanzhi Cheng, Zehao Li, Zheng Ma, Nuo Chen, Jialin Cao, Qiushi Sun, Zichen Ding, Fangzhi Xu, Hang Yan, Jiajun Chen, Anh Tuan Luu, Jianbing Zhang, Lewei Lu, Dahua Lin

  Paper Project Data Code

  An open recipe for mobile agents that synthesizes tasks from environment memory and rolls out trajectories with policy switching.
• ACL'26
  OS-Symphony: A Holistic Framework for Robust and Generalist Computer-Using Agents

  Bowen Yang, Kaiming Jin, Zhenyu Wu, Zhaoyang Liu, Qiushi Sun, Zehao Li, Jingjing Xie, Zhoumianze Liu, Fangzhi Xu, Kanzhi Cheng, Qingyun Li, Yian Wang, Yu Qiao, Zun Wang, Zichen Ding

  Paper

  Orchestrates tool, grounding, and reflection-memory agents for robust computer use across operating systems.
• ICLR'26 Oral
  ScaleCUA: Scaling Open-Source Computer Use Agents with Cross-Platform Data

  Zhaoyang Liu, Jingjing Xie, Zichen Ding, Zehao Li, Bowen Yang, Zhenyu Wu, Xuehui Wang, Qiushi Sun, Shi Liu, Weiyun Wang, Shenglong Ye, Qingyun Li, Xuan Dong, Yue Yu, Chenyu Lu, YunXiang Mo, Yao Yan, Zeyue Tian, Xiao Zhang, Yuan Huang, Yiqian Liu, Weijie Su, Gen Luo, Xiangyu Yue, Biqing Qi, Kai Chen, Bowen Zhou, Yu Qiao, Qifeng Chen, Wenhai Wang

  Paper HF Data Code

  Scales open computer-use agents with a cross-platform corpus spanning six operating systems.
• ICLR'25 Spotlight
  OS-ATLAS: A Foundation Action Model for Generalist GUI Agents

  Zhiyong Wu, Zhenyu Wu, Fangzhi Xu, Yian Wang, Qiushi Sun, Chengyou Jia, Kanzhi Cheng, Zichen Ding, Liheng Chen, Paul Pu Liang, Yu Qiao

  Paper Project HF Data ScreenSpot-v2

  A foundation action model unifying GUI grounding, action, and agent modes across platforms.
• ACL'24
  SeeClick: Harnessing GUI Grounding for Advanced Visual GUI Agents

  Kanzhi Cheng, Qiushi Sun, Yougang Chu, Fangzhi Xu, Yantao Li, Jianbing Zhang, Zhiyong Wu

  Paper Model Code

  Pioneers screenshot-only GUI agents via grounding pre-training and introduces the ScreenSpot benchmark.
• COLM'24
  Corex: Pushing the Boundaries of Complex Reasoning through Multi-Model Collaboration

  Qiushi Sun, Zhangyue Yin, Xiang Li, Zhiyong Wu, Xipeng Qiu, Lingpeng Kong

  Paper Slide Code BIB

  Multi-model collaboration that pushes complex reasoning beyond single-model prompting.
• Preprint
  OdysseyArena: Benchmarking Large Language Models For Long-Horizon, Active and Inductive Interactions

  Hang Yan^*, Fangzhi Xu^*, Qiushi Sun^*, Jinyang Wu, Zixian Huang, Muye Huang, Jingyang Gong, Zichen Ding, Kanzhi Cheng, Yian Wang, Xinyu Che, Zeyi Sun, Jian Zhang, Zhangyue Yin, Haoran Luo, Ben Kao, Qika Lin

  Paper Project

  Benchmarks LLMs on long-horizon, active, and inductive interaction in live environments.
• ACL'25
  Dynamic and Generalizable Process Reward Modeling

  Zhangyue Yin, Qiushi Sun, Zhiyuan Zeng, Qinyuan Cheng, Xipeng Qiu, Xuanjing Huang

  Paper

  Automatically designs process rewards and allocates them dynamically via reward trees and Pareto-optimal selection.
• ACL'25
  Interactive Evolution: A Neural-Symbolic Self-Training Framework for Large Language Models

  Fangzhi Xu, Qiushi Sun, Kanzhi Cheng, Jun Liu, Yu Qiao, Zhiyong Wu

  Paper

  A neural-symbolic self-training loop that improves LLMs from environment feedback without human annotation.
• ACL'24
  Boosting Language Models Reasoning with Chain-of-Knowledge Prompting

  Jianing Wang^*, Qiushi Sun^*, Xiang Li, Ming Gao

  Paper Slide

  Prompts models with structured knowledge triples to ground multi-step reasoning and curb hallucination.
• COLING'24
  Make Prompt-based Black-Box Tuning Colorful: Boosting Model Generalization from Three Orthogonal Perspectives

  Qiushi Sun, Chengcheng Han, Nuo Chen, Renyu Zhu, Jingyang Gong, Xiang Li, Ming Gao

  Paper

  Boosts gradient-free prompt tuning with two-stage optimizers, multi-verbalizers, and better initialization.
• Survey
  A Survey of Neural Code Intelligence: Paradigms, Advances and Beyond

  Qiushi Sun, Zhirui Chen, Fangzhi Xu, Chang Ma, Kanzhi Cheng, Zhangyue Yin, Jianing Wang, Chengcheng Han, Renyu Zhu, Shuai Yuan, Pengcheng Yin, Qipeng Guo, Xipeng Qiu, Xiaoli Li, Fei Yuan, Lingpeng Kong, Xiang Li, Zhiyong Wu

  Paper Slide Project Code BIB

  Traces the development of code intelligence, from pretrained models to LLM-based agents.