Papers by Zongmeng Zhang
Hybrid and Collaborative Passage Reranking (2023.findings-acl)
Copied to clipboard
| Challenge: | Existing solutions to passage reranking focus on enriching the interaction between query and each passage separately, neglecting the context among the top-ranked passages. |
| Approach: | They propose a Hybrid and Collaborative Passage Reranking method that leverages the similarity measurements of upstream retrievers for passage collaboration. |
| Outcome: | Experiments show that HybRank improves over existing methods and improves performance. |
BoolQuestions: Does Dense Retrieval Understand Boolean Logic in Language? (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Dense retrieval systems focus on optimizing text embedding space while overlooking Boolean logic in language. |
| Approach: | They propose a task to investigate whether retrieval systems can comprehend Boolean logic in language. |
| Outcome: | The proposed method is based on a benchmark dataset covering complex queries containing basic Boolean logic and corresponding annotated passages. |
Controllable Style Arithmetic with Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for linguistic style control lack fine-grained control, require extensive computation, or introduce significant latency. |
| Approach: | They propose a parameter-space approach that extracts style-specific representations by analyzing parameter differences between models trained on contrasting styles and incorporates them into a model with precise control over style intensity. |
| Outcome: | The proposed approach achieves three key capabilities while achieving optimal computational efficiency. |
Exploration-Exploitation Reshaping towards Efficient Reasoning for Large Language Models (2026.findings-acl)
Copied to clipboard
Yufeng Shi, Weilin Luo, Yuxiang Zhang, Zongmeng Zhang, Haoyang Liu, Yubing Wang, Bin Wang, Wengang Zhou, Houqiang Li
| Challenge: | Large Reasoning Models (LRMs) are constrained by the overthinking issue. |
| Approach: | They propose a policy optimization framework that reshapes the exploration and exploitation through two core components: self-imitation and self-guidance exploration. |
| Outcome: | The proposed model achieves superior reasoning efficiency without compromising overall accuracy. |