Latest news from Mingyue Cheng's research group at USTC.
2026
[May. 2026] π Congratulations to Xiaoyu Tao on our paper InstructTime++ being accepted to ACM TIST!
[May. 2026] π We released CastFactoryοΌηΌζεοΌ, a framework for LLM-driven time series forecasting model training. GitHub
[May. 2026] π Congratulations to Xiaoyu Tao and Bohou Zhang on our papers (TokenCast and ScholarSum) being accepted to IJCAI 2026!
[May. 2026] π Congratulations to Xiaoyu Tao and Yaguo Liu on our papers (CoGenCast and MemCast) being accepted to ICML 2026!
[Apr. 2026] π We released the preprint version of our survey on LLM-based agents from the contextual cognition perspective on Preprints.org: "A Comprehensive Survey of the LLM-Based Agent: The Contextual Cognition Perspective".
[Apr. 2026] π We released the preprint version of GeoMind on arXiv: "GeoMind: An Agentic Workflow for Lithology Classification with Reasoned Tool Invocation".
[Apr. 2026] π We released the preprint version of StepPO, our Agentic RL algorithm, on arXiv: "StepPO: Step-Aligned Policy Optimization for Agentic Reinforcement Learning".
[Apr. 2026] π We released the CastClawοΌθ§ζιοΌuser documentation, covering installation, configuration, and usage for agentic time series forecasting.
[Apr. 2026] π Congratulations to Jiawei Cao on his paper being accepted to ACL 2026 Main Conference!
[Apr. 2026] π Congratulations to Qingyang Mao on his paper being accepted to IEEE TKDE!
[Apr. 2026] π We released CastClawοΌθ§ζιοΌ, an open-source GitHub framework for agentic time series forecasting. Check out the project page!
[Apr. 2026] π We released Academic Search, an open-source GitHub skill for AI4Science literature search.
[Mar. 2026] π We open-sourced TabClaw, an GitHub framework for agentic table reasoning, designed to enable LLM agents to understand, analyze, and interact with large-scale tabular data through structured reasoning and tool-augmented workflows.
[Mar. 2026] π We open-sourced Claw-R1, an GitHub framework for agentic reinforcement learning, extending RL training environments for reasoning and tool-using LLM agents.
[Feb. 2026] π We released a Position Paper on arXiv: "Beyond Model-Centric Prediction β Agentic Time Series Forecasting".
[Jan. 2026] π We released PSPO, our Agentic RL algorithm, and validated it on the PaperScout paper search agent on arXiv: "PaperScout: An Autonomous Agent for Academic Paper Search with Process-Aware Sequence-Level Policy Optimization".
[Jan. 2026] π Congratulations to Li Li on her paper being accepted to DASFAA 2026!
[Jan. 2026] π Congratulations to Qingyang Mao and Zirui Liu on our paper being accepted to ICLR 2026!
[Jan. 2026] π I will serve as a Senior Program Committee (SPC) member for IJCAI2026.
[Jan. 2026] π Congratulations to Huajian Zhang, Shuo Yu, Qingchuan Li and Zhiding Liu on our paper being accepted to WWW 2026!
2025
[Dec. 2025] π I will serve as a AE for IEEE SMC2026.
[Nov. 2025] π Congratulations to Xiaoyu Tao on her paper being accepted to ACM TIST!
[Nov. 2025] π Congratulations to Qingchuan Li and Yupeng Li on our paper being accepted to AAAI 2026!
[Oct. 2025] π Congratulations to Chuang Jiang, Jiahao Wang, and collaborators on our paper being accepted to ACM WSDM2026!
[Sep. 2025] π Congratulations to Zhiding Liu on his paper being accepted to NeurIPS 2025!
[Aug. 2025] π Congratulations to Shuo Yu and Jiahao Wang on our paper being accepted to CIKM 2025!
[Jul. 2025] π Congratulations to Mingfan Pan and Yue Chen for our paper acceptance at RecSys 2025!
[Jun. 2025] π We released ChemTable, a large-scale dataset and benchmark for evaluating MLLMs on real-world chemical tables. [GitHub][arXiv]
[Jun. 2025] π We released Time-R1, the first time series reasoning model with reinforcement fine-tuning, now available on Hugging Face and GitHub.
[May. 2025] π Congratulations to Jie Ouyang and Junhao Yu on our paper being accepted to ACL 2025!
[May. 2025] π Congratulations to Daoyu Wang and Zirui Liu on our paper being accepted to ICML 2025!
[Apr. 2025] π Congratulations to Jintao Zhang, Yitong Zhou, and Hao Zhang for having our paper accepted by IJCAI 2025!
[Apr. 2025] π We preprinted a new survey: A Comprehensive Survey of Time Series Forecasting on TechRxiv.
[Apr. 2025] π Our survey A Survey on Table Mining with Large Language Models released on TechRxiv.
[Mar. 2025] π Our Agent R1 project officially released on GitHub.
[Mar. 2025] π Our survey A Survey on Knowledge-Oriented Retrieval-Augmented Generation released on arXiv.
2024
[Dec. 2024] π Congrats to our students on winning the Gold Award in the 2024 Ascend AI Innovation Competition finals!