Papers by Chenguang Wang
Want To Reduce Labeling Cost? GPT-3 Can Help (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Data annotation is labor-intensive and time-consuming for many NLP tasks. |
| Approach: | They propose to use GPT-3 to train models which are deployed for inference . they propose to combine pseudo labels from GPT3 with human labels . |
| Outcome: | The proposed method can be generalizable to many practical applications. |
ParaTag: A Dataset of Paraphrase Tagging for Fine-Grained Labels, NLG Evaluation, and Data Augmentation (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing datasets only annotate a binary label for each sentence pair. Existing models only annnotate binary labels for each phrase pair. |
| Approach: | They propose a novel binary paraphrase classification task that annotates the degree of paraphrase between sentences and a new annotation schema that labels the minimum spans of tokens in a sentence that don't have the corresponding paraphrases in the other sentence. |
| Outcome: | The proposed dataset can be used to train an automatic scorer for language generation evaluation. |
PEARL: Prompting Large Language Models to Plan and Execute Actions Over Long Documents (2024.eacl-long)
Copied to clipboard
| Challenge: | Using chain-of-thought prompting, large language models perform better on complex reasoning tasks. |
| Approach: | They propose a prompting framework that decomposes a question into a sequence of actions and executes them over the document to obtain the answer. |
| Outcome: | The proposed framework outperforms zero-shot and chain-of-thought prompting on a QuALITY dataset . it proposes a plan based on actions mined from a training set and executes it step by step . |
Retrieval Augmentation for Commonsense Reasoning: A Unified Approach (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for retrieving encyclopedic knowledge lack a large corpus and effective commonsense retriever. |
| Approach: | They propose a framework for retrieval-augmented commonsense reasoning with a large commonsensense corpus and a commonseense retriever. |
| Outcome: | The proposed framework outperforms existing methods on commonsense reasoning tasks. |
Fine-mixing: Mitigating Backdoors in Fine-tuned Language Models (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods for defending NLP models against backdoors have ignored the clean weights of PLMs. |
| Approach: | They exploit pre-trained weights to mitigate backdoors in fine-tuned NLP models . they use a fine-mixing technique and an Embedding Purification technique to do the same . |
| Outcome: | The proposed method outperforms baseline mitigation methods on three single-sentence sentiment classification tasks and two sentence-pair classification tasks. |
Fico: Evaluating Vision-Language Models under Visual Fidelity and Compression at Scale (2026.findings-acl)
Copied to clipboard
| Challenge: | Visual text compression is emerging paradigm for rendering text as images for processing by vision-language models. |
| Approach: | They propose a benchmark to assess VLM robustness under dense visual inputs. |
| Outcome: | Evaluating 13 general-purpose VLMs and 3 OCR-specialized models reveals performance drops sharply under increased density or reduced resolution; cross-task transfer between OCR, NIAH, and VQA is limited; and VQ is comparatively robust because low-level details are lost before high-level semantics. |
Zero-Shot Information Extraction as a Unified Text-to-Triple Translation (2021.emnlp-main)
Copied to clipboard
| Challenge: | a number of information extraction tasks require task-specific training. |
| Approach: | They propose a text-to-triple translation framework for information extraction tasks . they propose enabling task-agnostic translation by leveraging latent knowledge of a pre-trained language model . |
| Outcome: | The proposed framework outperforms the existing methods on open information extraction tasks. |
Z-Code++: A Pre-trained Language Model Optimized for Abstractive Summarization (2023.acl-long)
Copied to clipboard
Pengcheng He, Baolin Peng, Song Wang, Yang Liu, Ruochen Xu, Hany Hassan, Yu Shi, Chenguang Zhu, Wayne Xiong, Michael Zeng, Jianfeng Gao, Xuedong Huang
| Challenge: | Z-Code++ is a pre-trained language model optimized for abstractive text summarization. |
| Approach: | They propose a pre-trained language model optimized for abstractive text summarization that uses a two-phase pre-training technique to improve model's performance. |
| Outcome: | The proposed model outperforms the competing models on low-resource summarization tasks in zero-shot and few-shot settings. |
Benchmarking Language Models for Code Syntax Understanding (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Pre-trained language models capture the syntactic rules of natural languages without fine-tuning on syntax understanding tasks. |
| Approach: | They propose a benchmarking test to compare pre-trained language models with a large-scale dataset of programs annotated with syntactic relationships in their corresponding abstract syntax trees. |
| Outcome: | The proposed model fails to match baselines based on positional offsets and keywords. |
Empowering Language Models with Knowledge Graph Reasoning for Open-Domain Question Answering (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing Language Models lack the power to store all required knowledge, resulting in a lack of ability to infer out-of-context knowledge. |
| Approach: | They propose a Knowledge Interaction Layer that can be flexibly plugged into existing Transformer-based LMs to interact with a differentiable Knowledge Graph Reasoning module collaboratively. |
| Outcome: | The proposed model can be plugged into existing Transformer-based LMs to interact with a differentiable Knowledge Graph Reasoning module collaboratively. |
The Shifted and The Overlooked: A Task-oriented Investigation of User-GPT Interactions (2023.emnlp-main)
Copied to clipboard
Siru Ouyang, Shuohang Wang, Yang Liu, Ming Zhong, Yizhu Jiao, Dan Iter, Reid Pryzant, Chenguang Zhu, Heng Ji, Jiawei Han
| Challenge: | Recent advances in large language models (LLMs) have produced models that exhibit remarkable performance across a variety of NLP tasks. |
| Approach: | They analyze a large-scale collection of user-GPT conversations to identify a significant gap between academic research in NLP and the needs of real-world NLP applications. |
| Outcome: | The proposed model outperforms existing models in a large-scale collection of user-GPT conversations and identifies a significant gap between the tasks that users frequently request from LLMs and the tasks commonly studied in academic research. |
In-Context Demonstration Selection with Cross Entropy Difference (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Large language models (LLMs) can use in-context demonstrations to improve performance on zero-shot tasks. |
| Approach: | They propose a cross-entropy difference method for selecting in-context demonstrations that uses parameter efficient finetuning to train small models on training data. |
| Outcome: | The proposed method outperforms baseline selection methods on a mix-domain dataset and shows that the effectiveness of in-context demonstrations negatively correlates with the perplexity of the test example. |
PALT: Parameter-Lite Transfer of Language Models for Knowledge Graph Completion (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Pretrained language models (LMs) are a powerful transfer learning approach for knowledge graph (KG) completion. |
| Approach: | They propose a parameter-lite transfer learning approach for pretrained language models for knowledge graph (KG) completion. |
| Outcome: | The proposed model outperforms the state-of-the-art models on a knowledge graph completion benchmark by tuning 1% of the parameters. |
Retrieval Enhanced Model for Commonsense Generation (2021.findings-acl)
Copied to clipboard
| Challenge: | Existing frameworks for commonsense generation are lacking for pre-trained models. |
| Approach: | They propose a framework that uses concept matching to retrieve prototype sentences and trainable sentence retriever to enhance pre-training and fine-tuning. |
| Outcome: | The proposed framework achieves state-of-the-art on the large-scale Common-Gen benchmark. |
APOLLO: A Simple Approach for Adaptive Pretraining of Language Models for Logical Reasoning (2023.acl-long)
Copied to clipboard
Soumya Sanyal, Yichong Xu, Shuohang Wang, Ziyi Yang, Reid Pryzant, Wenhao Yu, Chenguang Zhu, Xiang Ren
| Challenge: | Existing methods to improve logical reasoning skills require complex data processing. |
| Approach: | They propose an adaptive pretraining approach to improve logical reasoning over text . they use a subset of Wikipedia sentences for pretraining and a sentence-level classification loss . |
| Outcome: | The proposed model outperforms baselines on LogiQA and ReClor. |
RuleR: Improving LLM Controllability by Rule-based Data Recycling (2025.naacl-short)
Copied to clipboard
| Challenge: | Existing supervised fine-tuning datasets are composed of general instructions without userspecified constraints. |
| Approach: | They propose a data augmentation method incorporating multiple constraints into the original data samples according to predefined rules to create new training tasks. |
| Outcome: | The proposed method improves LLM controllability while maintaining general instruction-following capabilities. |
AGENTVIGIL: Automatic Black-Box Red-teaming for Indirect Prompt Injection against LLM Agents (2025.findings-emnlp)
Copied to clipboard
Zhun Wang, Vincent Siu, Zhe Ye, Tianneng Shi, Yuzhou Nie, Xuandong Zhao, Chenguang Wang, Wenbo Guo, Dawn Song
| Challenge: | AGENTVIGIL is a black-box optimization framework to exploit indirect prompt injection vulnerabilities . indirect prompts compromise the core of LLM agents by manipulating contextual information rather than direct user prompts. |
| Approach: | They propose a black-box optimization framework to exploit indirect prompt injection vulnerabilities . they use a Monte Carlo tree-based algorithm to iteratively refine inputs . |
| Outcome: | The proposed framework achieves 71% and 70% success rates against two public benchmarks . |
Joint Language Semantic and Structure Embedding for Knowledge Graph Completion (2022.coling-1)
Copied to clipboard
| Challenge: | Existing methods to complete knowledge triplets rely on structures or semantics, but use semantics to improve performance. |
| Approach: | They propose to embed semantics in the natural language description of knowledge triplets with their structure information. |
| Outcome: | The proposed method improves performance on knowledge graph benchmarks and on low-resource regimes. |
Mosaic-IT: Cost-Free Compositional Data Synthesis for Instruction Tuning (2025.findings-acl)
Copied to clipboard
| Challenge: | Current instruction tuning relies on teacher models or human intervention to generate and refine the instructions and responses for training, which are costly, non-sustainable, and may lack diversity. |
| Approach: | They propose a human/model-free compositional data synthesis method that can create rich and diverse augmentations from existing instruction tuning data to enhance large language models. |
| Outcome: | The proposed method improves performance over benchmarks and reduces training costs by 80% compared with original instruction tuning. |
IELM: An Open Information Extraction Benchmark for Pre-Trained Language Models (2022.emnlp-main)
Copied to clipboard
| Challenge: | Recent studies show pre-trained LMs store linguistic and relational knowledge . pre-training LM models can answer "fill-in-the-blank" questions based on pre-defined relations . |
| Approach: | They propose an open information extraction benchmark for pre-trained language models . they turn pre-trained LMs into zero-shot OIE systems to examine open relational information . |
| Outcome: | The proposed benchmark outperforms state-of-the-art methods on factual OIE datasets without training sets. |
ProActor: Timing-Aware Reinforcement Learning for Proactive Task Scheduling Agents (2026.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to measuring and optimizing proactive task-oriented agents lack generalizable end-to-end solutions. |
| Approach: | They propose a framework for conversational task scheduling that integrates proactiveness reinforcement learning with a domain-agnostic annotation methodology. |
| Outcome: | The proposed framework enables scalable proactiveness reinforcement learning (RL) Experiments on two newly auto-annotated datasets demonstrate significant improvements in proactive timing while maintaining action consistency comparable to state-of-the-art baselines. |
Self-Generated Critiques Boost Reward Modeling for Language Models (2025.naacl-long)
Copied to clipboard
Yue Yu, Zhengxing Chen, Aston Zhang, Liang Tan, Chenguang Zhu, Richard Yuanzhe Pang, Yundi Qian, Xuewei Wang, Suchin Gururangan, Chao Zhang, Melanie Kambadur, Dhruv Mahajan, Rui Hou
| Challenge: | Existing reward models produce scalar scores and struggle to incorporate critiques in a natural language format. |
| Approach: | They propose a framework that predicts critiques and rewards using self-generated critiques without extra supervision. |
| Outcome: | The proposed framework improves reward modeling accuracy by 3.7%-7.3% compared to standard reward models and LLM judges. |
Leveraging Knowledge in Multilingual Commonsense Reasoning (2022.findings-acl)
Copied to clipboard
| Challenge: | Commonsense reasoning is a language-agnostic process, but most comprehensive knowledge sources are limited to a small number of languages, especially English. |
| Approach: | They propose to use English as a pivot language to integrate commonsense reasoning into models using a translate-retrieve-translate strategy. |
| Outcome: | The proposed model outperforms the state-of-the-art on the XCSR benchmarks. |
Auto-Instruct: Automatic Instruction Generation and Ranking for Black-Box Language Models (2023.findings-emnlp)
Copied to clipboard
Zhihan Zhang, Shuohang Wang, Wenhao Yu, Yichong Xu, Dan Iter, Qingkai Zeng, Yang Liu, Chenguang Zhu, Meng Jiang
| Challenge: | Large language models can perform a wide range of tasks by following natural language instructions without task-specific fine-tuning. |
| Approach: | They propose a method to automatically improve the quality of LLM instructions . they leverage the generative ability of LMS to generate diverse candidate instructions based on a scoring model trained on 575 existing NLP tasks. |
| Outcome: | The proposed method surpasses human-written and LLM-generated instructions on 118 out-of-domain tasks. |
Dict-BERT: Enhancing Language Model Pre-training with Dictionary (2022.findings-acl)
Copied to clipboard
Wenhao Yu, Chenguang Zhu, Yuwei Fang, Donghan Yu, Shuohang Wang, Yichong Xu, Michael Zeng, Meng Jiang
| Challenge: | Pre-trained language models (PLMs) capture word semantics in different contexts, hence the embeddings of rare words on the tail are poorly optimized. |
| Approach: | They propose to leverage definitions of rare words in dictionaries to enhance language model pre-training by leveraging dictionary definitions. |
| Outcome: | The proposed model improves understanding of rare words and boosts performance on various NLP downstream tasks. |
From Perceptions to Decisions: Wildfire Evacuation Decision Prediction with Behavioral Theory-informed LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing statistical methods for evacuation decision prediction fail to capture complex and diverse behavioral logic of different individuals. |
| Approach: | They propose a Large Language Model (LLM)-based framework that integrates behavioral theories and models to streamline the Chain-of-Thought reasoning and integrates with memory-based Reinforcement Learning module to provide accurate evacuation decision prediction and understanding. |
| Outcome: | The proposed framework improves on three post-wildfire survey datasets with strong cross-event generalizability over existing models. |
Re-Tuning: Overcoming the Compositionality Limits of Large Language Models with Recursive Tuning (2024.acl-long)
Copied to clipboard
| Challenge: | Existing methods to solve compositional tasks are limited by complexity and complexity. |
| Approach: | They propose a method that tunes large language models to break down a problem into subproblems, solve those subproblem, and combine the results. |
| Outcome: | The proposed method significantly improves model performance on three representative compositional tasks: integer addition, dynamic programming, and parity. |
AdaPrompt: Adaptive Model Training for Prompt-based NLP (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Prompt-based learning can tackle zero-shot and few-shot NLP tasks . authors propose a method that makes use of pre-trained language models . |
| Approach: | They propose to map NLP tasks into natural language prompts, which are then filled by pre-trained language models. |
| Outcome: | The proposed method outperforms standard prompt-based methods in few-shot settings. |
Small Models are Valuable Plug-ins for Large Language Models (2024.findings-acl)
Copied to clipboard
| Challenge: | Large-scale pre-trained language models are difficult to fine-tune due to their huge weights and limited context length. |
| Approach: | They propose an approach which allows black-box LLMs to work with locally fine-tuned smaller models, resulting in superior performance on supervised tasks. |
| Outcome: | The proposed approach overcomes the challenges of poor performance and instability of In-Context Learning (ICL) while reducing the complexity of in-context learning. |
InheritSumm: A General, Versatile and Compact Summarizer by Distilling from GPT (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Recent advances in large language models have revolutionized the way summarization is generated. |
| Approach: | They propose a summarization model derived from GPT-3.5 through distillation that is compact and has comparable summarizing capabilities to GPT-3. |
| Outcome: | The proposed model outperforms the established best small models in prefix-tuning and full-data fine-tuned scenarios. |
COSMIC: Generalized Refusal Direction Identification in LLM Activations (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing methods depend on predefined refusal templates detectable in output tokens or manual review. |
| Approach: | They propose a framework that optimally identifies steering directions and target layers using cosine similarity, entirely independent of output text. |
| Outcome: | The proposed framework achieves comparable steering effectiveness without any prior knowledge or assumptions of a model’s refusal behavior such as the use of certain refusal tokens. |
LMGQS: A Large-scale Dataset for Query-focused Summarization (2023.findings-emnlp)
Copied to clipboard
Ruochen Xu, Song Wang, Yang Liu, Shuohang Wang, Yichong Xu, Dan Iter, Pengcheng He, Chenguang Zhu, Michael Zeng
| Challenge: | Lack of large-scale datasets for query-focused summarization hinders model development . lack of data limits the ability of QFS models to train robust neural models . |
| Approach: | They propose to generate a query for each summary sentence in a generic summarization annotation using a pretrained language model. |
| Outcome: | The proposed model achieves state-of-the-art zero-shot and supervised performance on multiple existing QFS benchmarks. |
KG-FiD: Infusing Knowledge Graph in Fusion-in-Decoder for Open-Domain Question Answering (2022.acl-long)
Copied to clipboard
Donghan Yu, Chenguang Zhu, Yuwei Fang, Wenhao Yu, Shuohang Wang, Yichong Xu, Xiang Ren, Yiming Yang, Michael Zeng
| Challenge: | Open-Domain Question Answering (ODQA) models typically include a retrieving module and a reading module. |
| Approach: | They propose a new open-domain question-answering framework that uses a knowledge-enhanced version of FiD to improve the approach. |
| Outcome: | The proposed model improves on ODQA benchmark datasets with less than 40% computation cost. |
Task Compass: Scaling Multi-task Pre-training with Task Prefix (2022.findings-emnlp)
Copied to clipboard
Zhuosheng Zhang, Shuohang Wang, Yichong Xu, Yuwei Fang, Wenhao Yu, Yang Liu, Hai Zhao, Chenguang Zhu, Michael Zeng
| Challenge: | Existing studies show that multi-task learning with large-scale supervised tasks suffers from negative effects across tasks. |
| Approach: | They propose a task prefix guided multi-task pre-training framework to explore the relationships among tasks. |
| Outcome: | The proposed model can be used as a foundation backbone for a wide range of tasks and as augmentation tool for data augmentation with complementary tasks. |
Measuring Social Norms of Large Language Models (2024.findings-naacl)
Copied to clipboard
| Challenge: | Existing datasets that evaluate a general understanding of social science are inadequate to understand social norms. |
| Approach: | They propose a multi-agent framework to improve large language models’ ability to understand social norms by comparing them to elementary students. |
| Outcome: | The proposed framework improves large language models to be on par with humans. |
DeepStruct: Pretraining of Language Models for Structure Prediction (2022.findings-acl)
Copied to clipboard
| Challenge: | Pretrained language models perform structural understanding tasks that focus on understanding one aspect of the text. |
| Approach: | They propose a method for improving the structural understanding abilities of language models by pretraining them to generate structures from the text on task-agnostic corpora. |
| Outcome: | The proposed model performs state-of-the-art on 21 of 28 datasets. |
G-Eval: NLG Evaluation using Gpt-4 with Better Human Alignment (2023.emnlp-main)
Copied to clipboard
| Challenge: | Conventional reference-based metrics have low correlation with human judgments, especially for open-ended generation tasks. |
| Approach: | They propose to use large language models as reference-free NLG evaluators to assess the quality of NLG outputs. |
| Outcome: | The proposed framework outperforms all previous methods in two generation tasks, and has a Spearman correlation of 0.514 with human on summarization task, and a large variance in human judgments. |
Training Data is More Valuable than You Think: A Simple and Effective Method by Retrieving from Training Data (2022.acl-long)
Copied to clipboard
| Challenge: | Experimental results show that REtrieving from the traINing datA only can lead to significant gains on multiple NLG and NLU tasks. |
| Approach: | They propose to retrieve training instances from traINing datA and concatenate them with input to generate output. |
| Outcome: | The proposed method achieves state-of-the-art results on XSum, BigPatent, and CommonsenseQA. |
Ensuring Readability and Data-fidelity using Head-modifier Templates in Deep Type Description Generation (P19-1)
Copied to clipboard
| Challenge: | Existing generative methods overlook grammatical structure or make factual mistakes in generated texts. |
| Approach: | They propose a template-based method to ensure the readability of generated type descriptions . they also propose measurable metrics to measure the readibility of the generated type description . |
| Outcome: | The proposed method improves substantially compared with baselines and achieves state-of-the-art performance on both datasets. |
PoD: Positional Dependency-Based Word Embedding for Aspect Term Extraction (2020.coling-main)
Copied to clipboard
| Challenge: | Existing word embeddings that capture the contextual information only produce moderate results in aspect term extraction. |
| Approach: | They propose a positional dependency-based word embedding which takes both dependency context and positional context into account for aspect term extraction. |
| Outcome: | The proposed method outperforms other embedding methods in aspect term extraction. |