HERO: Hindsight-Enhanced Reflection from Environment Observations for Agentic Self-Distillation

Haoran Liu*, Yuwei Zhang*, Xiyao Li, Bohan Lyu, Jingbo Shang (* equal contribution)

Preprint 2026

HERO turns next environment observations into hindsight reflection signals for multi-turn agent self-distillation, improving task success and reducing unnecessary tool-use turns under limited training budgets.

[Paper]

CoMem: Context Management with A Decoupled Long-Context Model

Yuwei Zhang, Chengyu Dong, Shuowei Jin, Changlong Yu, Hejie Cui, Hongye Jin, Xinyang Zhang, Hamed Bonab, Colin Lockard, Jianshu Chen, Zhenyu Shi, Jingbo Shang, Xian Li, Bing Yin

ICML 2026

CoMem decouples agent reasoning from memory summarization so long-horizon agents can preserve most long-context performance while reducing latency through asynchronous context management.

[Paper] [Code]

ChipMATE: Multi-Agent Training via Reinforcement Learning for Enhanced RTL Generation

Zhongkai Yu, Yichen Lin, Chenyang Zhou, Yuwei Zhang, Kun Zhou, Junxia Cui, Haotian Ye, Zhengding Hu, Zaifeng Pan, Ruiyi Wang, Yujie Zhao, Hejia Zhang, Jingbo Shang, Jishen Zhao, Yufei Ding

Preprint 2026

ChipMATE trains Verilog and Python reference-model agents to mutually verify RTL generation without a golden oracle, improving deployable hardware-code generation for industrial settings.

[Paper] [Code]

Learning with Rare Success but Rich Feedback via Reflection-Enhanced Self-Distillation

Yuwei Zhang, Sha Li, Changlong Yu, Qin Lu, Shuowei Jin, Chengyu Dong, Haoran Liu, Ilgee Hong, Xintong Li, Zhenyu Shi, Bing Yin, Jingbo Shang

Preprint 2026

RESD turns failed rollouts into reflection-based supervision and reusable playbook knowledge, letting models improve efficiently even when successful rollouts are rare.

[Paper] [Project Page] [Code]

MaPPO: Maximum a posteriori preference optimization with prior knowledge

Guangchen Lan, Sipeng Zhang, Tianle Wang, Yuwei Zhang, Daoan Zhang, Xinpeng Wei, Xiaoman Pan, Hongming Zhang, Dong-Jun Han, Christopher G Brinton

Preprint 2025

MaPPO reframes preference optimization as a maximum-a-posteriori objective that injects prior reward knowledge to improve alignment without extra hyperparameters.

[Paper]

Attention Reveals More Than Tokens: Training-Free Long-Context Reasoning with Attention-guided Retrieval

Yuwei Zhang, Jayanth Srinivasa, Gaowen Liu, Jingbo Shang

Preprint 2025

Attrieval uses attention from chain-of-thought tokens as a training-free signal to retrieve implicit facts that long-context LLMs miss during multi-step reasoning.

[Paper]

Toward Multi-Session Personalized Conversation: A Large-Scale Dataset and Hierarchical Tree Framework for Implicit Reasoning

Xintong Li, Jalend Bantupalli, Ria Dharmani, Yuwei Zhang, Jingbo Shang

EMNLP 2025 Oral

ImplexConv and TaciTree target implicit personalization across many conversation sessions by structuring memory into a hierarchy that supports efficient, fine-grained reasoning.

[Paper]

Speculative RAG: Enhancing Retrieval Augmented Generation through Drafting

Zilong Wang, Zifeng Wang, Long Le, Steven Zheng, Swaroop Mishra, Vincent Perot, Yuwei Zhang, Anush Mattapalli, Ankur Taly, Jingbo Shang, Chen-Yu Lee, Tomas Pfister

ICLR 2025

Speculative RAG speeds up retrieval-augmented generation by letting a small specialist draft evidence-grounded answers in parallel while a larger model verifies them once.

[Paper]

Controllable Data Augmentation for Few-Shot Text Mining with Chain-of-Thought Attribute Manipulation

Letian Peng, Yuwei Zhang, Jingbo Shang

ACL Findings 2024

[Paper]

Answer is All You Need: Instruction-following Text Embedding via Answering the Question

Letian Peng*, Yuwei Zhang*, Zilong Wang, Jayanth Srinivasa, Gaowen Liu, Zihan Wang, Jingbo Shang (* equal contribution)

ACL 2024

[Paper] [Code]

ClusterLLM: Large Language Models as a Guide for Text Clustering

Yuwei Zhang, Zihan Wang, Jingbo Shang

EMNLP 2023

[Paper] [Code]

Fine-tuning Pre-trained Language Models for Few-shot Intent Detection: Supervised Pretraining and Isotropization

Haode Zhang, Haowen Liang, Yuwei Zhang, Liming Zhan, Xiao-Ming Wu, Xiaolei Lu, Albert Y.S. Lam

NAACL 2022 Oral

[Paper] [Code]

New Intent Discovery with Pre-training and Contrastive Learning

Yuwei Zhang, Haode Zhang, Li-Ming Zhan, Xiao-Ming Wu, Albert Y.S. Lam

ACL 2022

[Paper] [Code]

Development and validation of a screening model for lung cancer using machine learning: a large-scale, multi-center study of biomarkers in breath

Jing Li*, Yuwei Zhang*, Qing Chen, Zhenhua Pan, Jun Chen, Meixiu Sun, Junfeng Wang, Yingxin Li, Qing Ye (* equal contribution)

Frontiers in Oncology 2022

[Paper]

Automated Calculation of Liquid Crystal Sensing Images Based on Deep Learning

Yuwei Zhang*, Shuai Xu*, Ronghua Zhang, Zhichao Deng, Yin Liu, Jianguo Tian, Li Yu, Qiongzheng Hu, Qing Ye (* equal contribution)

Analytical Chemistry 2022

[Paper]

Effectiveness of Pre-training for Few-shot Intent Classification

Haode Zhang*, Yuwei Zhang*, Li-Ming Zhan, Jiaxin Chen, Guangyuan Shi, Xiao-Ming Wu, Albert Y.S. Lam (* equal contribution)

EMNLP Findings 2021

[Paper] [Code]