Papers by Zhen Zeng
CRITICTOOL: Evaluating Self-Critique Capabilities of Large Language Models in Tool-Calling Error Scenarios (2025.emnlp-main)
Copied to clipboard
| Challenge: | a number of tools are used to perform complex tasks, but the tool utilization process can cause errors. |
| Approach: | They propose a critique evaluation benchmark for tool learning that analyzes function-calling errors on tool evaluation benchmarks. |
| Outcome: | The proposed critique evaluation benchmark holds diverse tool-use errors with varying complexities, which better reflects real-world scenarios. |
LAW: Legal Agentic Workflows for Custody and Fund Services Contracts (2025.coling-industry)
Copied to clipboard
William Watson, Nicole Cho, Nishan Srishankar, Zhen Zeng, Lucas Cecchi, Daniel Scott, Suchetha Siddagangappa, Rachneet Kaur, Tucker Balch, Manuela Veloso
| Challenge: | Currently, there are limited resources available to build a legal domain-specific Large Language Model (LLM) however, legal contracts are highly varied not only in terms of semantics but also accessibility. |
| Approach: | They propose a Large Language Model (LLM) that integrates multiple specialized agents and text agents to respond to user queries. |
| Outcome: | The proposed model outperforms the baseline model in complex tasks such as calculating a contract’s termination date by 92.9% points. |
UltraEval-Audio: A Unified Framework for Comprehensive Evaluation of Audio Foundation Models (2026.acl-demo)
Copied to clipboard
Qundong Shi, Jie Zhou, Biyuan Lin, Junbo Cui, Guoyang Zeng, Yixuan Zhou, Ziyang Wang, Xin Liu, Zhen Luo, Yudong Wang, Zhiyuan Liu
| Challenge: | Existing evaluation frameworks for audio foundation models are heavily reliant on English, making it difficult to objectively assess models’ performance on Chinese. |
| Approach: | They propose a unified framework that supports 10 languages, 14 task categories, 24 models, and 36 benchmarks with one-command evaluation and real-time leaderboards. |
| Outcome: | The proposed framework supports 10 languages, 14 task categories, 24 models, and 36 benchmarks with one-command evaluation and real-time leaderboards. |
MM-BizRAG: Rethinking Multimodal Retrieval-Augmented Generation for General Purpose Enterprise Q&A (2026.acl-industry)
Copied to clipboard
Hanoz Bhathena, Parin Rajesh Jhaveri, Rohan Mittal, Prateek Singh, Aymen Kallala, Rachneet Kaur, Yiqiao Jin, Zhen Zeng, Adwait Ratnaparkhi, Denis Kochedykov
| Challenge: | Recent advances in multimodal retrieval-augmented generation (MM-RAG) have shifted toward minimal parsing, relying on page-level images for producing retriever embeddings and answer generation. |
| Approach: | They propose a document structure-aware split that extracts and represents document structure via a structure-based split that dynamically routes documents through orientation-specific ingestion pipelines. |
| Outcome: | The proposed model outperforms state-of-the-art vision-centric baselines by up to 32% points and achieves strong gains on report-style layouts. |
DCP: Dual-Cue Pruning for Efficient Large Vision-Language Models (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing pruning methods for large vision language models use visual tokens to prune . existing methods fail to balance efficiency and semantic alignment due to large number of visual token. |
| Approach: | They propose a cross-modal pruning framework that considers textual semantics and visual self-attention to combine them to achieve efficient inference acceleration. |
| Outcome: | The proposed pruning framework can retain only 25% of the visual tokens, with a minimal performance degradation of only 0.063% on LLaVA-1.5-13B. |
SlideAgent: Hierarchical Agentic Framework for Multi-Page Visual Document Understanding (2026.acl-long)
Copied to clipboard
| Challenge: | Multimodal large language models (MLLMs) are a promising tool for document understanding, but they are not able to handle complex multi-page visual documents. |
| Approach: | They propose a flexible agentic framework for understanding multi-modal, multi-page, and multi-layout documents . SlideAgent employs specialized agents and decomposes reasoning into three specialized levels . |
| Outcome: | a new agentic framework improves accuracy over open-source and proprietary models . it decomposes reasoning into three levels to capture themes and visual cues . the framework is based on a multimodal large language model and a MLLM . |
Stepwise Perplexity-Guided Refinement for Efficient Chain-of-Thought Reasoning in Large Language Models (2025.findings-acl)
Copied to clipboard
Yingqian Cui, Pengfei He, Jingying Zeng, Hui Liu, Xianfeng Tang, Zhenwei Dai, Yan Han, Chen Luo, Jing Huang, Zhen Li, Suhang Wang, Yue Xing, Jiliang Tang, Qi He
| Challenge: | Chain-of-Thought (CoT) reasoning has improved the performance of large language models (LLMs) however, the detailed reasoning process in CoT often incurs long generation times and high computational costs due to the inclusion of unnecessary steps. |
| Approach: | They propose a method to identify critical reasoning steps using perplexity as a measure of their importance. |
| Outcome: | The proposed method achieves a better balance between reasoning accuracy and efficiency of CoT. |
CLEEK: A Chinese Long-text Corpus for Entity Linking (2020.lrec-1)
Copied to clipboard
| Challenge: | Entity linking is a fundamental task in natural language processing, says nigel kilgstrom . existing corpora for entity linking in china are lacking and deficient, he says . kilsmstrom: a new method for entity disambiguation can be developed for Chinese . |
| Approach: | They build a Chinese corpus of multi-domain long text for entity linking . they evaluate the difficulty of documents with respect to entity linking using a measure . |
| Outcome: | The proposed corpus is based on 100 documents from diverse domains and is publicly accessible. |
AdaptAgent: Adapting Multimodal Web Agents with Few-Shot Learning from Human Demonstrations (2025.acl-long)
Copied to clipboard
| Challenge: | State-of-the-art multimodal web agents can perform many web tasks by processing user instructions and interacting with graphical user interfaces (GUIs). |
| Approach: | They propose to build multimodal web agents for few-shot adaptability using human demonstrations to improve their generalization and adaptability. |
| Outcome: | The proposed framework enables both proprietary and open-weights multimodal web agents to adapt to new websites and domains using few human demonstrations. |
Evaluating Large Language Models on Time Series Feature Understanding: A Comprehensive Taxonomy and Benchmark (2024.emnlp-main)
Copied to clipboard
Elizabeth Fons, Rachneet Kaur, Soham Palande, Zhen Zeng, Tucker Balch, Manuela Veloso, Svitlana Vyetrenko
| Challenge: | Large Language Models (LLMs) are a critical tool for time series analysis and reporting in many fields, including healthcare, finance, climate, and many more. |
| Approach: | They propose a framework for rigorously evaluating the capabilities of Large Language Models (LLMs) on time series understanding, encompassing both univariate and multivariate forms. |
| Outcome: | The proposed framework delineates various characteristics inherent in time series data. |
A General Framework to Enhance Fine-tuning-based LLM Unlearning (2025.findings-acl)
Copied to clipboard
Jie Ren, Zhenwei Dai, Xianfeng Tang, Hui Liu, Jingying Zeng, Zhen Li, Rahul Goutam, Suhang Wang, Yue Xing, Qi He, Hui Liu
| Challenge: | Existing approaches to remove copyrighted and privacy-sensitive data from Large Language Models (LLMs) have been proposed to remove specific data from LLMs without requiring full retraining. |
| Approach: | They propose a general framework that enhances the utility of fine-tuning-based methods by distinguishing target data and suppressing related generations. |
| Outcome: | The proposed framework improves the unlearning and utility of fine-tuning-based methods by distinguishing the target data and suppressing related generations. |
LETS-C: Leveraging Text Embedding for Time Series Classification (2025.acl-long)
Copied to clipboard
| Challenge: | Recent advances in language modeling have shown promising results when applied to time series data. |
| Approach: | They propose a method to fine-tune large language models for time series classification tasks using text embedding models and a simple classification head. |
| Outcome: | The proposed model outperforms the current SOTA model on a time series classification benchmark and uses only 14.5% of the trainable parameters. |
ChartAgent: A Multimodal Agent for Visually Grounded Reasoning in Complex Chart Question Answering (2026.acl-long)
Copied to clipboard
| Challenge: | Recent multimodal LLMs have shown promise in chart-based visual question answering, but their performance declines sharply on unannotated charts. |
| Approach: | They propose a novel agentic framework that explicitly performs visual reasoning directly within the chart’s spatial domain. |
| Outcome: | The proposed framework achieves state-of-the-art accuracy on the ChartBench and ChartX benchmarks surpassing prior methods by up to 16.07% absolute gain overall and 17.31% on numerically intensive queries. |
RAGLAB: A Modular and Research-Oriented Unified Framework for Retrieval-Augmented Generation (2024.emnlp-demo)
Copied to clipboard
Xuanwang Zhang, Yun-Ze Song, Yidong Wang, Shuyun Tang, Xinfeng Li, Zhengran Zeng, Zhen Wu, Wei Ye, Wenyuan Xu, Yue Zhang, Xinyu Dai, Shikun Zhang, Qingsong Wen
| Challenge: | Existing research on Retrieval Augmented Generation (RAG) does not address the problem of hallucinations and real-time updating of knowledge. |
| Approach: | They propose a modular open-source library to equip LLMs with external knowledge. |
| Outcome: | The proposed approach reduces the need for expensive open-source tools and lacks fair comparisons between novel RAG algorithms. |
UniCorn: Towards Self-Improving Unified Multimodal Models through Self-Generated Supervision (2026.acl-long)
Copied to clipboard
Zhen Fang, Ruiyan Han, XinYu Sun, Yuchen Ma, Ziheng Wang, Yu Zeng, Zehui Chen, Lin Chen, Wenxuan Huang, Wei-Jie Xu, Yi Cao, Feng Zhao
| Challenge: | Unified Multimodal Models have achieved remarkable success in cross-modal comprehension, but a gap persists in their ability to translate internal knowledge into faithful and controllable synthesis. |
| Approach: | They propose a self-improvement framework that partitions a single UMM into three collaborative roles: Proposer, Solver, and Judge. |
| Outcome: | The proposed framework improves on TIIF, DPG, CompBench and UniCycle benchmarks. |
Unveiling Privacy Risks in LLM Agent Memory (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Model (LLM) agents store private user-agent interactions in memory for demonstrations, introducing new privacy risks for LLM agents. |
| Approach: | They propose an attack that extracts private information from memory under a black-box setting and propose a method that can be used to attack the agent. |
| Outcome: | The proposed attack is effective under a black-box setting and it is demonstrated on two representative agents. |