Papers by Zifeng Wang
Doc2SoarGraph: Discrete Reasoning over Visually-Rich Table-Text Documents via Semantic-Oriented Hierarchical Graphs (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing work on document visual question answering fails to capture the differences and correlations between elements of a document and associated questions. |
| Approach: | They propose a document-visual question-answering challenge that exploits element-level semantics and employs hierarchical Graph structures to capture differences and correlations between elements. |
| Outcome: | The proposed model surpasses the state-of-the-art method and large language model in terms of Exact Match (EM) metric, demonstrating exceptional effectiveness. |
Magnet: Multi-turn Tool-use Data Synthesis and Distillation via Graph Translation (2025.acl-long)
Copied to clipboard
Fan Yin, Zifeng Wang, I-Hung Hsu, Jun Yan, Ke Jiang, Yanfei Chen, Jindong Gu, Long Le, Kai-Wei Chang, Chen-Yu Lee, Hamid Palangi, Tomas Pfister
| Challenge: | Large language models have been shown to be effective in multi-turn interactions . however, their performance may be limited in complex, multi-turned interactions involving users and multiple tools. |
| Approach: | They propose a framework for synthesizing high-quality training trajectories to enhance the function calling capability of large language model agents in multi-turn conversations with humans. |
| Outcome: | The proposed model outperforms the teacher model by 68.01 on BFCL-v3 and 73.30 on ToolQuery. |
QueryForm: A Simple Zero-shot Form Entity Query Framework (2023.findings-acl)
Copied to clipboard
Zifeng Wang, Zizhao Zhang, Jacob Devlin, Chen-Yu Lee, Guolong Su, Hao Zhang, Jennifer Dy, Vincent Perot, Tomas Pfister
| Challenge: | Form-like document understanding is a key yet under-investigated problem . endlessly training specialized models on new document types is not scalable in many practical scenarios. |
| Approach: | They propose to use large-scale query-entity pairs generated from form-like webpages to pre-train QueryForm. |
| Outcome: | The proposed framework sets state-of-the-art average F1 score on XFUND and Payment benchmarks. |
Token Prepending: A Training-Free Approach for Eliciting Better Sentence Embeddings from LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Recent studies have focused on prompt engineering to extract sentence embeddings from large language models (LLMs) but these models are mostly decoder-only and the earlier tokens in the sentence cannot attend to the latter, resulting in biased encoding of sentence information and cascading effects on the final decoded token. |
| Approach: | They propose a plug-and-play and training-free technique that prepends each layer’s decoded sentence embedding to the beginning of the sentence in the next layer’ s input. |
| Outcome: | The proposed technique can significantly improve the performance of existing prompt-based sentence embedding methods across different LLMs while incurring negligible additional inference cost. |
SAGE: Steerable Agentic Data Generation for Deep Search with Execution Feedback (2026.findings-eacl)
Copied to clipboard
Fangyuan Xu, Rujun Han, Yanfei Chen, Zifeng Wang, I-Hung Hsu, Jun Yan, Vishy Tirumalashetty, Eunsol Choi, Tomas Pfister, Chen-Yu Lee
| Challenge: | High-quality, complex question-answer pairs are pivotal for training and evaluating capable deep search agents. |
| Approach: | They propose a pipeline that generates high-quality, difficulty-controlled deep search question-answer pairs for a given corpus and a target difficulty level. |
| Outcome: | The proposed pipeline generates high-quality, difficulty-controlled deep search question-answer pairs for a given corpus and a target difficulty level. |
SciAssess: Benchmarking LLM Proficiency in Scientific Literature Analysis (2025.findings-naacl)
Copied to clipboard
Hengxing Cai, Xiaochen Cai, Junhan Chang, Sihang Li, Lin Yao, Wang Changxin, Zhifeng Gao, Hongshuai Wang, Li Yongge, Mujie Lin, Shuwen Yang, Jiankun Wang, Mingjun Xu, Jin Huang, Xi Fang, Jiaxi Zhuang, Yuqi Yin, Yaqi Li, Changhong Chen, Zheng Cheng, Zifeng Zhao, Linfeng Zhang, Guolin Ke
| Challenge: | Existing benchmarks fail to adequately evaluate the proficiency of Large Language Models (LLMs) Existing standards do not cover the skills needed to evaluate LLMs in scientific literature analysis. |
| Approach: | They propose a benchmark to evaluate the proficiency of large language models in scientific literature analysis. |
| Outcome: | SciAssess evaluates 11 LLMs on multiple tasks across scientific fields. |
Aggregating Multiple Heuristic Signals as Supervision for Unsupervised Automated Essay Scoring (2023.acl-long)
Copied to clipboard
| Challenge: | Automated Essay Scoring (AES) aims to evaluate the quality score of input essays without human intervention. |
| Approach: | They propose an unsupervised approach to evaluate the quality of input essays . they use multiple heuristic quality signals as pseudo-groundtruths to train a neural AES model . |
| Outcome: | The proposed approach achieves state-of-the-art performance on eight prompts of ASPA dataset compared with previous unsupervised methods . |
MindMap: Knowledge Graph Prompting Sparks Graph of Thoughts in Large Language Models (2024.acl-long)
Copied to clipboard
| Challenge: | Large language models suffer from limitations such as difficulty in incorporating new knowledge, generating hallucinations, and explaining their reasoning process. |
| Approach: | They propose a pipeline that leverages knowledge graphs to enhance LLMs’ inference and transparency by eliciting the mind map of LLM's, which reveals their reasoning pathways based on the ontology of knowledge. |
| Outcome: | The proposed pipeline enables LLMs to comprehend KG inputs and infer with a combination of implicit and external knowledge. |
s3: You Don’t Need That Much Data to Train a Search Agent via RL (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing approaches to optimize retrieval using search-only metrics ignore downstream utility and fine-tune entire LLM to jointly reason and retrieve limit retrieval utility and compatibility with frozen or proprietary models. |
| Approach: | They propose a lightweight, model-agnostic framework that decouples the searcher from the generator and trains the search user using a Gain Beyond RAG reward. |
| Outcome: | The proposed framework outperforms baselines trained on over 70 more data with 2.4k training samples. |
Focusing Condition: Inference-Time Self-Contrastive Steering Elicits Better Conditional Text Embeddings in LLMs (2026.acl-long)
Copied to clipboard
| Challenge: | Existing methods for extracting conditional text embeddings from large language models (LLMs) relying on prompts often fails to produce high-quality conditional embeddables, resulting in degradation of quality. |
| Approach: | They propose a plug-and-play method that constructs unconditional general text embeddings and uses them to refine conditional text embeds. |
| Outcome: | The proposed method improves performance of prompt-based methods on clustering, Semantic Textual Similarity, and triplet alignment datasets. |
Reverse Thinking Makes LLMs Stronger Reasoners (2025.naacl-long)
Copied to clipboard
Justin Chen, Zifeng Wang, Hamid Palangi, Rujun Han, Sayna Ebrahimi, Long Le, Vincent Perot, Swaroop Mishra, Mohit Bansal, Chen-Yu Lee, Tomas Pfister
| Challenge: | Reverse-Enhanced Thinking (RevThink) is a framework for large language models to perform reverse thinking. |
| Approach: | They propose a framework for enhancing forward-backward reasoning by collecting data from a teacher model and employing three objectives to train a student model in a multi-task learning fashion. |
| Outcome: | The proposed framework outperforms a fine-tuning method trained on 10x more forward reasoning on 12 datasets covering commonsense, math, and logical reasoning. |
Found in the middle: Calibrating Positional Attention Bias Improves Long Context Utilization (2024.findings-acl)
Copied to clipboard
Cheng-Yu Hsieh, Yung-Sung Chuang, Chun-Liang Li, Zifeng Wang, Long Le, Abhishek Kumar, James Glass, Alexander Ratner, Chen-Yu Lee, Ranjay Krishna, Tomas Pfister
| Challenge: | Large language models struggle to capture relevant information located in the middle of their input. |
| Approach: | They propose a calibration mechanism that allows the model to attend to contexts faithfully according to their relevance even when they are in the middle. |
| Outcome: | The proposed calibration mechanism mitigates this positional bias and improves retrieval-augmented generation performance. |
In Prospect and Retrospect: Reflective Memory Management for Long-term Personalized Dialogue Agents (2025.acl-long)
Copied to clipboard
Zhen Tan, Jun Yan, I-Hung Hsu, Rujun Han, Zifeng Wang, Long Le, Yiwen Song, Yanfei Chen, Hamid Palangi, George Lee, Anand Rajan Iyer, Tianlong Chen, Huan Liu, Chen-Yu Lee, Tomas Pfister
| Challenge: | Existing approaches to long-term dialogue memory management fail to capture the natural semantic structure of conversations, leading to fragmented and incomplete representations. |
| Approach: | They propose a mechanism that integrates forward- and backward-looking reflections into a personalized memory bank for effective future retrieval. |
| Outcome: | The proposed mechanism outperforms state-of-the-art benchmarks on a long-term dialogue memory model. |
PlanGEN: A Multi-Agent Framework for Generating Planning and Reasoning Trajectories for Complex Problem Solving (2025.emnlp-main)
Copied to clipboard
Mihir Parmar, Xin Liu, Palash Goyal, Yanfei Chen, Long Le, Swaroop Mishra, Hossein Mobahi, Jindong Gu, Zifeng Wang, Hootan Nakhost, Chitta Baral, Chen-Yu Lee, Tomas Pfister, Hamid Palangi
| Challenge: | Existing methods for natural planning lack constraint-guided iterative verification and adaptive selection . a recent study found that LLMs are not good at such planning. |
| Approach: | They propose a model-agnostic and easily scalable agent framework with three key components: constraint, verification, and selection agents. |
| Outcome: | The proposed framework improves inference-time algorithms on NATURAL PLAN and OlympiadBench benchmarks. |
TriSum: Learning Summarization Ability from Large Language Models with Structured Rationale (2024.naacl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have advanced tasks like text summarization, but their size and computational demands limit their use in resource-constrained and privacy-centric settings. |
| Approach: | They propose a framework for distilling LLMs’ text summarization abilities into a compact, local model using a curriculum learning strategy that evolves from simple to complex tasks. |
| Outcome: | The proposed framework outperforms baseline models on CNN/DailyMail, XSum, and ClinicalTrial, and improves interpretability by providing insights into the summarization rationale. |
GenRES: Rethinking Evaluation for Generative Relation Extraction in the Era of Large Language Models (2024.naacl-long)
Copied to clipboard
| Challenge: | Existing relation extraction methods rely on exact matching with human-annotated reference relations, while GRE methods produce diverse and semantically accurate relations. |
| Approach: | They propose a multi-dimensional assessment of relation extraction methods using human-annotated reference relations. |
| Outcome: | The proposed method is consistent with human preferences for RE quality. |
AutoTrial: Prompting Language Models for Clinical Trial Design (2023.emnlp-main)
Copied to clipboard
| Challenge: | Generative large language models (LLMs) are a popular tool for creating coherent and human-like documents for clinical trials. |
| Approach: | They propose to generate clinical eligibility criteria using language models by a hybrid of discrete and neural prompting and scalable knowledge incorporation via in-context learning. |
| Outcome: | The proposed method generates high-quality criteria texts fluent and coherent with high accuracy against the GPT-3.5 baselines. |
MedCLIP: Contrastive Learning from Unpaired Medical Images and Text (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing vision-text contrastive learning methods encounter many false negatives, i.e., images and reports from separate patients probably carry the same semantics but are wrongly treated as negatives. |
| Approach: | They propose to decouple medical image-text contrastive learning and replace it with semantic matching loss based on medical knowledge to eliminate false negatives in contrastive training. |
| Outcome: | The proposed framework outperforms state-of-the-art methods on zero-shot prediction, supervised classification, and image-text retrieval with only 20K pre-training data. |
PILOT: Legal Case Outcome Prediction with Case Law (2024.naacl-long)
Copied to clipboard
| Challenge: | predicting legal case outcomes requires identifying relevant precedent cases . predicting case outcomes in case law systems presents unique challenges . |
| Approach: | They propose a framework for making legal case outcome predictions with case law . they propose to use two modules for relevant case retrieval and temporal pattern handling . |
| Outcome: | The proposed framework shows significant improvement over previous models based on civil law cases . it is crucial to identify relevant precedent cases that serve as evidence for judges . |
CaLM: Contrasting Large and Small Language Models to Verify Grounded Generation (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing methods to generate grounded responses are prone to errors due to the irrelevancy of input documents. |
| Approach: | They propose a framework that leverages the insight that a robust grounded response should be consistent with information derived solely from its cited sources. |
| Outcome: | Experiments on three open-domain question-answering datasets show that the proposed framework improves performance by 1.5% to 7% without any model fine-tuning. |
Finding Influential Instances for Distantly Supervised Relation Extraction (2022.coling-1)
Copied to clipboard
| Challenge: | Distant supervision models suffer from high label noise and are not reliable for DS. |
| Approach: | They propose a model-agnostic instance sampling method for relation extraction (RE) by influence function, namely REIF. |
| Outcome: | The proposed method reduces the computational complexity from O(mn) to O(1), with analyzing its robustness on the selected sampling function. |
KEEP CHATTING! An Attractive Dataset for Continuous Conversation Agents (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing works about persona dialogue such as PersonaChat have greatly facilitated the chatbot with configurable and persistent personalities. |
| Approach: | They propose to collect a dataset called ContinuousChat and rewrite it in style-specific ways to increase users' willingness to continue chatting. |
| Outcome: | The proposed model increases users' willingness to continue talking to the chatbot by increasing their personas to detailed-personas through experiences, daily life, future plans, or interesting stories. |
BMIKE-53: Investigating Cross-Lingual Knowledge Editing with In-Context Learning (2025.acl-long)
Copied to clipboard
| Challenge: | Using a benchmark for cross-lingual knowledge editing, knowledge editing is underexplored. |
| Approach: | They propose a benchmark for cross-lingual in-context knowledge editing that spans 53 languages and three KE datasets. |
| Outcome: | The proposed benchmark systematically evaluates cross-lingual knowledge editing (IKE) under zero-shot, one-shot and few-shot setups. |
Contrastive Prompting Enhances Sentence Embeddings in LLMs through Inference-Time Steering (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies focus on prompt engineering to encode the full semantics of a sentence into the embedding of the last token. |
| Approach: | They propose a technique that introduces an extra auxiliary prompt to elicit better sentence embedding . they propose to use the hidden state of the token as the sentence embedded in LLMs . |
| Outcome: | The proposed technique can improve performance of existing prompt-based methods on STS tasks and downstream classification tasks. |
PromptEHR: Conditional Electronic Healthcare Records Generation with Prompt Learning (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for generating longitudinal multimodal EHRs are limited due to privacy concerns. |
| Approach: | They propose to generate longitudinal multimodal EHRs by unconditional generation or longitudinal inference . existing methods generate single-modal E HRs by conditional generation or by longitudinal inferment . |
| Outcome: | The proposed method is more flexible and controllable than existing methods and is more cost-effective than existing ones. |
When One LLM Drools, Multi-LLM Collaboration Rules (2026.acl-long)
Copied to clipboard
Shangbin Feng, Wenxuan Ding, Alisa Liu, Zifeng Wang, Weijia Shi, Yike Wang, Shannon Zejiang Shen, Xiaochuang Han, Hunter Lang, Chen-Yu Lee, Tomas Pfister, Yejin Choi, Yulia Tsvetkov
| Challenge: | a single general-purpose LLM is not enough to produce a reliable output, argues this paper . a multi-LLM collaboration approach addresses reliability, democratization, and pluralism . |
| Approach: | They argue that a single general-purpose LLM is not enough to produce a reliable output . they organize existing multi-LLM collaboration methods into a hierarchy based on access and information exchange . |
| Outcome: | The proposed method addresses reliability, democratization, and pluralism challenges a single LLM fails to produce a reliable output. |
Trial2Vec: Zero-Shot Clinical Trial Document Similarity Search using Self-Supervision (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Clinical trials are expensive and time-consuming to conduct, and lengthy trial documents and lack of labeled data make comparisons difficult. |
| Approach: | They propose a zero-shot clinical trial retrieval method which learns through self-supervision without the need for annotating similar clinical trials. |
| Outcome: | The proposed method improves on baselines on precision/recall and 15% on the downstream trial outcome prediction task. |
AEA: Adaptive Expert Allocation Improves Sentence Embeddings from Mixture-of-Experts LLM (2026.acl-long)
Copied to clipboard
| Challenge: | Existing methods to improve embeddings from Mixture-of-Experts models allocate a fixed number of experts uniformly across all layers and tokens, ignoring inter-layer and inter-token heterogeneity. |
| Approach: | They propose an Adaptive Expert Allocation framework that performs layer-wise and token-wise expert allocation to enhance embedding quality. |
| Outcome: | The proposed method improves embedding quality across multiple MoE models. |
ProtoCycle: Reflective Tool-Augmented Planning for Text-Guided Protein Design (2026.findings-acl)
Copied to clipboard
Yutang Ge, Guojiang Zhao, Sihang Li, Zheng Cheng, Zifeng Zhao, Hanchen Xia, Guolin Ke, Linfeng Zhang, Zhifeng Gao, Yu Guang Wang
| Challenge: | Recent deep generative models have already shown encouraging * Equal contribution. |
| Approach: | They propose to use generic instruction-tuned LLMs as direct text-to-sequence generators to achieve this goal. |
| Outcome: | Recent studies show that reflection improves sequence quality and alignment while maintaining competitive foldability. |
Multi-Prompting Decoder Helps Better Language Understanding (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing methods to adapt Pre-trained Language Models to downstream tasks are limited by their inference APIs. |
| Approach: | They propose a multi-prompting decoding framework that query PLMs with multiple prompts . they propose to query Plms with optimal transport for hidden states and calibrated decoding for class scores . |
| Outcome: | The proposed framework achieves state-of-the-art results on multiple natural language understanding datasets under the few-shot setting. |
LMDX: Language Model-based Document Information Extraction and Localization (2024.findings-acl)
Copied to clipboard
Vincent Perot, Kai Kang, Florian Luisier, Guolong Su, Xiaoyu Sun, Ramya Sree Boppana, Zilong Wang, Zifeng Wang, Jiaqi Mu, Hao Zhang, Chen-Yu Lee, Nan Hua
| Challenge: | Large Language Models have revolutionized Natural Language Processing but their application in extracting information from visually rich documents has not been successful. |
| Approach: | They propose a language model-based document information extraction and localization methodology to reframe the document information extract task for a LLM. |
| Outcome: | The proposed method enables extraction of singular, repeated, and hierarchical entities with and without training data. |
CodecLM: Aligning Language Models with Tailored Synthetic Data (2024.findings-naacl)
Copied to clipboard
Zifeng Wang, Chun-Liang Li, Vincent Perot, Long Le, Jin Miao, Zizhao Zhang, Chen-Yu Lee, Tomas Pfister
| Challenge: | Recent work on generating diverse instructions and applying LLM to increase instruction complexity neglects downstream use cases. |
| Approach: | They propose a framework for generating high-quality synthetic data for LLM alignment with different downstream instruction distributions and LLMs. |
| Outcome: | Experiments on four open-domain instruction using the proposed framework validate the effectiveness of CodecLM over the current state-of-the-art. |