Papers by Hua Wu

CORD: Bridging the Audio–Text Reasoning Gap via Weighted On-policy Cross-modal Distillation (2026.findings-acl)

Challenge:	Existing studies rely on additional syntax-driven attention components to enhance the transformer, which require more parameters and additional syntactic parsing in downstream tasks.
Approach:	They propose a syntax-guided contrastive learning method which does not change the transformer architecture and does not alter the transformer structure.
Outcome:	The proposed method achieves consistent improvements in a variety of tasks including grammatical error detection, entity tasks, structural probing and GLUE.

Copied to clipboard

Hu Jing, Danxiang Zhu, Xianlong Luo, Dan Zhang, Shuwei He, Yishu Lei, Shikun Feng, Hai-Tao Zheng, Jingzhou HE, Yu Sun, Hua Wu, Haifeng Wang

Learning In-context Learning for Named Entity Recognition (2023.acl-long)

Challenge:	Large Audio Language Models (LALMs) exhibit a degradation in knowledge and reasoning capabilities . empirical results show that CORD significantly bridges the audio–text performance gap .
Approach:	They propose a framework that performs online cross-modal self-distillation to bridge the acoustic-semantic gap between LALMs and text-based models.
Outcome:	The proposed framework bridges the acoustic-semantic gap between LALMs and text-based models . it employs on-policy reverse KL divergence with importance-aware weighting .

Copied to clipboard

Jiawei Chen, Yaojie Lu, Hongyu Lin, Jie Lou, Wei Jia, Dai Dai, Hua Wu, Boxi Cao, Xianpei Han, Le Sun

DuReader_robust: A Chinese Dataset Towards Evaluating Robustness and Generalization of Machine Reading Comprehension in Real-World Applications (2021.acl-short)

Challenge:	Existing methods to recognize entities in text are limited by the diversity of entity types and the lack of high-quality annotations.
Approach:	They propose an in-context learning-based NER approach that can inject in-const NER ability into PLMs and recognize entities of novel types on-the-fly using only a few demonstrative instances.
Outcome:	The proposed method outperforms the PLMs+fine-tuning counterparts on 4 few-shot NER datasets and significantly outperformed the Plms+initialized extractors.

Copied to clipboard

Hongxuan Tang, Hongyu Li, Jing Liu, Yu Hong, Hua Wu, Haifeng Wang

Addressing Troublesome Words in Neural Machine Translation (D18-1)

Challenge:	In order to comprehensively verify the robustness and generalization of MRC models, we construct a real-world Chinese dataset - DuReader_robust .
Approach:	They introduce a real-world Chinese dataset to evaluate the robustness and generalization of MRC models from three aspects: over-sensitivity, over-stability and generalisation.
Outcome:	The proposed model fails to perform well on the challenge test set and may provide suggestions for future model development.

Copied to clipboard

Yang Zhao, Jiajun Zhang, Zhongjun He, Chengqing Zong, Hua Wu

Machine Translation

Challenge:	Neural machine translation (NMT) has weaknesses in handling lowfrequency and ambiguous words, which we refer to as troublesome words.
Approach:	They propose to use contextual memory to memorize which target words should be produced in which situations to translate troublesome words.
Outcome:	The proposed method outperforms baseline models on Chinese-to-English and English-to German translation tasks.

Uncertainty-Aware Routing for Principled Alignment with MoE Dynamics (2026.acl-long)

Copied to clipboard

Yilong Chen, Junyuan Shang, Yuchen Feng, Zhenyu Zhang, Naibin Gu, Ziqi Wang, Tingwen Liu, Shuohuan Wang, Yu Sun, Hua Wu, Haifeng Wang

Challenge:	Mixture-of-Experts (MoE) is a cornerstone for scaling LLMs, yet its training dynamics remain poorly understood, often leading to sub-optimal specialization.
Approach:	They propose to use Helmholtz Free Energy and Router Entropy to study the MoE lifecycle and identify a universal Three-Stage Phase Transition .
Outcome:	The proposed model reduces perplexity and improves expert distinctiveness, offering a principled path toward thermodynamically aligned computation.

SKEP: Sentiment Knowledge Enhanced Pre-training for Sentiment Analysis (2020.acl-main)

Copied to clipboard

Hao Tian, Can Gao, Xinyan Xiao, Hao Liu, Bolei He, Hua Wu, Haifeng Wang, Feng Wu

Q-TOD: A Query-driven Task-oriented Dialogue System (2022.emnlp-main)

Challenge:	sentiment knowledge is ignored in sentiment analysis, despite its use in pretraining.
Approach:	They propose to use sentiment knowledge to learn a unified sentiment representation for multiple sentiment analysis tasks.
Outcome:	The proposed method outperforms strong pre-training baseline on three kinds of sentiment tasks.

Copied to clipboard

Xin Tian, Yingzhan Lin, Mengfei Song, Siqi Bao, Fan Wang, Huang He, Shuqi Sun, Hua Wu

Distributional Clarity: The Hidden Driver of RL-Friendliness in Large Language Models (2026.acl-long)

Challenge:	Existing pipelined task-oriented dialogue systems have difficulties adapting to unseen domains . end-to-end systems are plagued by large-scale knowledge bases in practice .
Approach:	They propose a query-driven task-oriented dialogue system that extracts dialogue context information into a natural language query.
Outcome:	The proposed system outperforms strong baselines and establishes a new state-of-the-art performance on three publicly available task-oriented dialogue datasets.

Copied to clipboard

Shaoning Sun, Mingzhu Cai, Huang He, Bingjin Chen, Siqi Bao, Yujiu Yang, Hua Wu, Haifeng Wang

TOME: A Two-stage Approach for Model-based Retrieval (2023.acl-long)

Challenge:	RL-friendly models exhibit intra-class compactness and inter-class separation in probability assignments . under identical training, Qwen models achieve substantial gains, while others like Llama yield limited improvements.
Approach:	They propose a method to quantify distributional clarity in probability space . they show distributional clearness is a trainable property underlying RL-Friendliness .
Outcome:	The proposed model families achieve substantial gains under identical training, while others like Llama yield limited improvements.

Copied to clipboard

Ruiyang Ren, Wayne Xin Zhao, Jing Liu, Hua Wu, Ji-Rong Wen, Haifeng Wang

PLATO-Ad: A Unified Advertisement Text Generation Framework with Multi-Task Prompt Learning (2022.emnlp-industry)

Challenge:	Recent research has focused on model-based retrieval, which discards the index in the traditional retrieval model and memorizes the candidate corpora using model parameters.
Approach:	They propose a model-based retrieval approach that discards the index in the traditional retrieval model and memorizes the candidate corpora using model parameters.
Outcome:	The proposed approach eliminates the index in the traditional retrieval model and memorizes the candidate corpora using model parameters.

Copied to clipboard

Zeyang Lei, Chao Zhang, Xinchao Xu, Wenquan Wu, Zheng-yu Niu, Hua Wu, Haifeng Wang, Yi Yang, Shuanglong Li

Unified Structure Generation for Universal Information Extraction (2022.acl-long)

Challenge:	Online advertisement text generation models have achieved remarkable success in generating high-quality text ads, but some challenges remain, such as low-resource scenarios and training efficiency for multiple ad tasks.
Approach:	They propose a unified text ad generation framework with multi-task prompt learning to tackle low-resource ade generation problem and a multi-step prompt learning mechanism to efficiently solve multiple aed generation tasks.
Outcome:	The proposed framework outperforms the state-of-the-art on offline and online metrics.

Copied to clipboard

Yaojie Lu, Qing Liu, Dai Dai, Xinyan Xiao, Hongyu Lin, Xianpei Han, Le Sun, Hua Wu

Conversational Graph Grounded Policy Learning for Open-Domain Conversation Generation (2020.acl-main)

Challenge:	Information extraction suffers from its varying targets, heterogeneous structures, and demand-specific schemas.
Approach:	They propose a unified text-to-structure generation framework, namely UIE, which can universally model different IE tasks, adaptively generate targeted structures, and collaboratively learn general IE abilities from different knowledge sources.
Outcome:	The proposed framework can model different IE tasks, generate targeted structures, and learn general IE abilities from different knowledge sources.

Copied to clipboard

Jun Xu, Haifeng Wang, Zheng-Yu Niu, Hua Wu, Wanxiang Che, Ting Liu

Learning Adaptive Segmentation Policy for Simultaneous Translation (2020.emnlp-main)

Challenge:	Existing word-level policy models that learn dialog policy and language generation from dialog corpora often lead to degeneration issues where the utterances become ungrammatical or repetitive.
Approach:	They propose to represent prior dialog transitions as a graph and learn a CG grounded dialog policy that can foster a more coherent and controllable dialog.
Outcome:	The proposed framework is able to learn dialog policy in open-domain multi-turn conversation.

Copied to clipboard

Ruiqing Zhang, Chuanqiang Zhang, Zhongjun He, Hua Wu, Haifeng Wang

Text Generation

Challenge:	Experimental results show that adaptive segmentation policies for simultaneous translation are more accurate than current methods . if translation starts before adequate source content is delivered, the quality of translation degrades . waiting for too much source text increases latency, which would hurt accuracy .
Approach:	They propose a new adaptive segmentation policy for simultaneous translation based on human interpreters . it learns to segment the source text by considering possible translations produced by the translation model .
Outcome:	Experimental results show that the proposed method achieves better accuracy-latency trade-off over state-of-the-art methods.

Unveiling Environmental Impacts of Large Language Model Serving: A Functional Unit View (2025.acl-long)

Copied to clipboard

Yanran Wu, Inez Hua, Yi Ding

Multi-Turn Response Selection for Chatbots with Deep Attention Matching Network (P18-1)

Challenge:	Existing studies benchmark carbon emissions but lack a standardized basis for comparison across different model configurations.
Approach:	They propose a functional unit (FU) as a standardized basis for comparison across different model configurations and introduce FUEL, the first FU-based framework for evaluating LLM serving’s environmental impact.
Outcome:	The proposed framework uncovers key insights and trade-offs in reducing carbon emissions by optimizing model size, quantization strategy, and hardware choice, paving the way for more sustainable LLM serving.

Copied to clipboard

Xiangyang Zhou, Lu Li, Daxiang Dong, Yi Liu, Ying Chen, Wayne Xin Zhao, Dianhai Yu, Hua Wu

UNIMO: Towards Unified-Modal Understanding and Generation via Cross-Modal Contrastive Learning (2021.acl-long)

Challenge:	Existing models for matching dialogue responses rely on semantic and functional dependencies . a recent study only uses the last utterance in context for matching a reply .
Approach:	They propose a model that matches a response with its multi-turn context using attention.
Outcome:	The proposed model outperforms the state-of-the-art models on two large-scale multi-turn response selection tasks.

Copied to clipboard

Wei Li, Can Gao, Guocheng Niu, Xinyan Xiao, Hao Liu, Jiachen Liu, Hua Wu, Haifeng Wang

DuReadervis: A Chinese Dataset for Open-domain Document Visual Question Answering (2022.findings-acl)

Challenge:	Existing pre-training methods focus on single-modal tasks or multi-modal ones . large-scale pre- training has drawn much attention in both the community of Compute Vision (CV) and Natural Language Processing (NLP).
Approach:	They propose a UNIfied-MOdal pre-training architecture which can adapt to both single-modal and multi-modal understanding and generation tasks.
Outcome:	The proposed model can learn more generalizable representations with rich non-paired single-modal data.

Copied to clipboard

Le Qi, Shangwen Lv, Hongyu Li, Jing Liu, Yu Zhang, Qiaoqiao She, Hua Wu, Haifeng Wang, Ting Liu

STACL: Simultaneous Translation with Implicit Anticipation and Controllable Latency using Prefix-to-Prefix Framework (P19-1)

Challenge:	Open-domain question answering is a task that requires answering questions based on a collection of document images.
Approach:	They propose to use document images to answer questions using layouts and visual features instead of text.
Outcome:	The proposed approach reduces human cost and improves scalability of QA systems by incorporating layouts and visual features.

Copied to clipboard

Mingbo Ma, Liang Huang, Hao Xiong, Renjie Zheng, Kaibo Liu, Baigong Zheng, Chuanqiang Zhang, Zhongjun He, Hairong Liu, Xing Li, Hua Wu, Haifeng Wang

LEMON: Reviving Stronger and Smaller LMs from Larger LMs with Linear Parameter Fusion (2024.acl-long)

Challenge:	Simultaneous translation is notoriously dif- ficult due to word-order differences.
Approach:	They propose a prefix-to-prefix framework that implicitly learns to anticipate in a single translation model.
Outcome:	The proposed framework achieves low latency and reasonable qual- ity on 4 directions.

Copied to clipboard

Yilong Chen, Junyuan Shang, Zhenyu Zhang, Shiyao Cui, Tingwen Liu, Shuohuan Wang, Yu Sun, Hua Wu

DuReader-Retrieval: A Large-scale Chinese Benchmark for Passage Retrieval from Web Search Engine (2022.emnlp-main)

Challenge:	Existing methods to train a stronger and smaller model with the help of large models are limited by the model size and performance.
Approach:	They propose to learn competent initial points for smaller models by fusing parameters from larger models and introduce controllable receptive fields to model prior parameter characteristics.
Outcome:	The proposed method outperforms baselines in terms of effectiveness and efficiency.

Copied to clipboard

Yifu Qiu, Hongyu Li, Yingqi Qu, Ying Chen, QiaoQiao She, Jing Liu, Hua Wu, Haifeng Wang

AttnPO: Attention-Guided Process Supervision for Efficient Reasoning (2026.acl-long)

Challenge:	Existing datasets for non-English passage retrieval are lacking in quality and accuracy.
Approach:	They present a large-scale Chinese dataset for passage retrieval . they reduce false negatives by manually annotating results pooled from multiple retrievers .
Outcome:	The proposed dataset reduces false negatives in development and testing sets and removes similar training queries.

Copied to clipboard

Shuaiyi Nie, null Dingsiyu, Wenyuan Zhang, Linhao Yu, Tianmeng Yang, Yao Chen, Weichong Yin, Yu Sun, Hua Wu, Tingwen Liu

Clip-Tuning: Towards Derivative-free Prompt Learning with a Mixture of Rewards (2022.findings-emnlp)

Challenge:	Existing trajectory-level length penalties fail to effectively shorten reasoning length and degrade accuracy, as they treat all reasoning steps uniformly and lack fine-grained signals to distinguish redundancy from necessity.
Approach:	They propose a low-overhead process-supervised RL framework that leverages the model’s intrinsic attention signals for step-level credit assignment.
Outcome:	The proposed framework reduces reasoning length while improving performance across 9 benchmarks.

Copied to clipboard

Yekun Chai, Shuohuan Wang, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang

Learning Adaptive Segmentation Policy for End-to-End Simultaneous Translation (2022.acl-long)

Challenge:	Existing work does not take full advantage of over-parameterized characteristics of large pre-trained language models.
Approach:	They propose a method that uses frozen "thinned" networks to obtain a mixture of rewards and advance the derivative-free prompt learning.
Outcome:	The proposed method outperforms previous gradient-free prompt learning methods and achieves parity with gradient-based counterparts on seven language understanding benchmarks under few-shot settings.

Copied to clipboard

Ruiqing Zhang, Zhongjun He, Hua Wu, Haifeng Wang

Speech & Audio in NLP

Challenge:	Existing methods to perform simultaneous speech-to-text translation ignore contextual information and suffer from low translation quality.
Approach:	They propose an adaptive segmentation policy for simultaneous speech-to-text translation . it learns to segment the source streaming speech into meaningful units .
Outcome:	The proposed method achieves a good accuracy-latency trade-off over state-of-the-art methods on English-German and Chinese-English.

Discovering Dialog Structure Graph for Coherent Dialog Generation (2021.acl-long)

Copied to clipboard

Jun Xu, Zeyang Lei, Haifeng Wang, Zheng-Yu Niu, Hua Wu, Wanxiang Che

FRSUM: Towards Faithful Abstractive Summarization via Enhancing Factual Robustness (2022.findings-emnlp)

Challenge:	Existing studies on dialog structure graphs from open-domain dialogs have limited number of dialog states and can be laborious and costly to annotate manually.
Approach:	They propose to use dialog structure graph as a model to discover hierarchical latent dialog states and their transitions from corpus to facilitate dialog management in a RL based dialog system.
Outcome:	The proposed model can discover meaningful dialog structure graph and significantly improve multi-turn coherence on two benchmark corpora.

Copied to clipboard

Wenhao Wu, Wei Li, Jiachen Liu, Xinyan Xiao, Ziqiang Cao, Sujian Li, Hua Wu

Robustness in NLP

Challenge:	Existing models of abstractive summarization are able to generate fluent and coherent summaries, but they still suffer from the unfaithful generation problem.
Approach:	They propose to improve the faithfulness of existing models by enhancing their factual robustness by using a novel training strategy, namely FRSUM, which teaches the model to defend against both explicit adversarial samples and implicit factual adversarials.
Outcome:	The proposed training strategy improves faithfulness of various models, such as T5, BART, and T5 .

Where to Go for the Holidays: Towards Mixed-Type Dialogs for Clarification of User Goals (2022.acl-long)

Copied to clipboard

Zeming Liu, Jun Xu, Zeyang Lei, Haifeng Wang, Zheng-Yu Niu, Hua Wu

PLATO-XL: Exploring the Large-scale Pre-training of Dialogue Generation (2022.findings-aacl)

Challenge:	a dialog system posits that users have figured out clear and specific goals . but in many real-world scenarios, users struggle to figure out specific goals by determining all the necessary slots.
Approach:	They propose a mixed-type dialog model with a Prompt-based continual learning mechanism . they collect 5k dialog sessions and 168k utterances for 4 dialog types and 5 domains .
Outcome:	The proposed model provides user-goal-related knowledge to help figure out clear and specific goals . it can be extended to any specific type by utilizing existing dialog corpora effectively.

Copied to clipboard

Siqi Bao, Huang He, Fan Wang, Hua Wu, Haifeng Wang, Wenquan Wu, Zhihua Wu, Zhen Guo, Hua Lu, Xinxian Huang, Xin Tian, Xinchao Xu, Yingzhan Lin, Zheng-Yu Niu

Dynamic Prefix as Instructor for Incremental Named Entity Recognition: A Unified Seq2Seq Generation Framework (2025.findings-acl)

Challenge:	Experimental results show PLATO-XL achieves state-of-the-art results across multiple conversational tasks.
Approach:	They propose to train PLATO-XL models with up to 11 billion parameters, trained on Chinese and English social media conversations.
Outcome:	The proposed model achieves state-of-the-art on multiple conversational tasks, verifying its potential as a foundation model of conversational AI.

Copied to clipboard

Zihao Wu, YongXiang Hua, Yongxin Zhu, Fang Zhang, Linli Xu

PAIR: Leveraging Passage-Centric Similarity Relation for Improving Dense Passage Retrieval (2021.findings-acl)

Challenge:	Named Entity Recognition (NER) is a fundamental problem in information extraction.
Approach:	They propose a parameter-efficient method for Incremental Named Entity Recognition (INER) task aimed at updating a model to extract entities from an expanding set of entity type candidates by employing a dynamic prefix as a task instructor to guide the generative model.
Outcome:	Empirical results show that the proposed method preserves task-invariant knowledge while adapting to new entities with minimal parameter updates.

Copied to clipboard

Ruiyang Ren, Shangwen Lv, Yingqi Qu, Jing Liu, Wayne Xin Zhao, QiaoQiao She, Hua Wu, Haifeng Wang, Ji-Rong Wen

Proactive Human-Machine Conversation with Explicit Conversation Goal (P19-1)

Challenge:	Recent studies only consider query-centric similarity relation when learning the dual-encoder retriever.
Approach:	They propose a query-centric and PAssage-centric approach to capture more comprehensive similarity relations for dense passage retrieval.
Outcome:	The proposed approach significantly outperforms existing models on both MSMARCO and Natural Questions datasets.

Copied to clipboard

Wenquan Wu, Zhen Guo, Xiangyang Zhou, Hua Wu, Xiyuan Zhang, Rongzhong Lian, Haifeng Wang

QDMR-based Planning-and-Solving Prompting for Complex Reasoning Tasks (2024.lrec-main)

Challenge:	Typical human-machine conversation systems only use utterances and responses as training data, which results in uninformative and inappropriate responses.
Approach:	They propose a dataset where one acts as a conversation leader and the other as 'follower' they establish baseline results on a 270K utterances and 30k dialogues dataset using state-of-the-art models.
Outcome:	The proposed model can generate diverse multi-turn conversations using knowledge from a new dataset .

Copied to clipboard

Jinfeng Huang, Qiaoqiao She, Wenbin Jiang, Hua Wu, Yang Hao, Tong Xu, Feng Wu

Guardian-as-an-Advisor: Advancing Next-Generation Guardian Models for Trustworthy LLMs (2026.findings-acl)

Challenge:	Existing Plan-and-Solve prompting methods are difficult to implement for complex questions.
Approach:	They propose a plan-and-solve prompting method based on Question Decomposition Meaning Representation (QDMR) it allows LLM to generate a QDMR graph to represent problem-solving logic .
Outcome:	The proposed method can represent and execute the problem-solving logic of complex questions more accurately than existing methods.

Copied to clipboard

Yue Huang, Haomin Zhuang, Jiayi Ye, Han Bao, Yanbo Wang, Hang Hua, Siyuan Wu, Pin-Yu Chen, Xiangliang Zhang

ERNIE-M: Enhanced Multilingual Representation by Aligning Cross-lingual Semantics with Monolingual Corpora (2021.emnlp-main)

Challenge:	prevailing taxonomies neglect robustness and honesty, yielding safer-on-paper but less useful systems.
Approach:	They propose a soft-gating pipeline where a guardian predicts a binary risk label plus a concise explanation and prepends this advice to the original query for re-inference.
Outcome:	The proposed model maintains safety while reducing over-refusal.

Copied to clipboard

Xuan Ouyang, Shuohuan Wang, Chao Pang, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang

Inner Thinking Transformer: Leveraging Dynamic Depth Scaling to Foster Adaptive Internal Thinking (2025.acl-long)

Challenge:	Existing methods for pretraining cross-lingual models are limited in their size due to the limited amount of parallel corpora.
Approach:	They propose a method that encourages the model to align multiple languages with monolingual corpora to overcome the constraint of the parallel corpus size.
Outcome:	The proposed method outperforms existing cross-lingual models and delivers new state-of-the-art results in various cross-linguistic downstream tasks.

Copied to clipboard

Yilong Chen, Junyuan Shang, Zhenyu Zhang, Yanxi Xie, Jiawei Sheng, Tingwen Liu, Shuohuan Wang, Yu Sun, Hua Wu, Haifeng Wang

Leveraging Graph to Improve Abstractive Multi-Document Summarization (2020.acl-main)

Challenge:	Large language models face inherent performance bottlenecks under parameter constraints . challenging tokens induce abrupt gradient spikes across layers, exposing stress points .
Approach:	They propose an inner thinking transformer that reimagines layer computations as implicit thinking steps.
Outcome:	Empirical results show that ITT outperforms Transformer/Loop variants in 11 benchmarks.

Copied to clipboard

Wei Li, Xinyan Xiao, Jiachen Liu, Hua Wu, Haifeng Wang, Junping Du

NACL: A General and Effective KV Cache Eviction Framework for LLM at Inference Time (2024.acl-long)

Challenge:	Empirical results show that our model brings substantial improvements over several strong baselines.
Approach:	They propose a neural abstractive multi-document summarization model which captures cross-document relations and can guide the summary generation process.
Outcome:	The proposed model improves on the WikiSum and MultiNews datasets and can be easily combined with pre-trained language models.

Copied to clipboard

Yilong Chen, Guoxia Wang, Junyuan Shang, Shiyao Cui, Zhenyu Zhang, Tingwen Liu, Shuohuan Wang, Yu Sun, Dianhai Yu, Hua Wu

Diversified Multiple Instance Learning for Document-Level Multi-Aspect Sentiment Classification (2020.emnlp-main)

Challenge:	Large Language Models (LLMs) with extended context windows are expensive and infeasible on fixed memory hardware due to the surprisingly large memory consumption of KV Cache.
Approach:	They propose a general framework for long-context KV cache eviction that achieves more optimal and efficient evict in a single operation during the encoding phase.
Outcome:	The proposed framework improves performance on short- and long-text tasks by 80% and 76% respectively, reducing KV Cache by up to 5 with over 95% performance maintenance.

Copied to clipboard

Yunjie Ji, Hao Liu, Bolei He, Xinyan Xiao, Hua Wu, Yanhua Yu

Enhancing Local Feature Extraction with Global Representation for Neural Text Classification (D19-1)

Challenge:	Experimental results show that D-MILN outperforms recent weakly-supervised baselines . document-level multi-aspect sentiment classification requires a lot of manual aspect-level annotations - which is time-consuming and laborious .
Approach:	They propose a novel Diversified Multiple Instance Learning Network to achieve DMSC with only document-level weak supervision.
Outcome:	The proposed method outperforms weakly-supervised baselines on TripAdvisor and BeerAdvocate datasets.

Copied to clipboard

Guocheng Niu, Hengru Xu, Bolei He, Xinyan Xiao, Hua Wu, Sheng Gao

CDConv: A Benchmark for Contradiction Detection in Chinese Conversations (2022.emnlp-main)

Challenge:	Existing methods for text classification learn long dependency by deeply stacking or hybrid modeling.
Approach:	They propose a global-based local feature extraction architecture with global information incorporated into the local feature extractor.
Outcome:	The proposed architecture outperforms the previous best models on eight benchmark datasets.

Copied to clipboard

Chujie Zheng, Jinfeng Zhou, Yinhe Zheng, Libiao Peng, Zhen Guo, Wenquan Wu, Zheng-Yu Niu, Hua Wu, Minlie Huang

Multi-agent Learning for Neural Machine Translation (D19-1)

Challenge:	Existing methods for detecting dialogue contradictions are difficult due to contextualization nature of conversations.
Approach:	They propose a benchmark for Contradiction Detection in Chinese Conversations . they use automatic conversation generation to simulate common user behaviors .
Outcome:	The proposed benchmark simulated the user behaviors that trigger chatbots to make contradictions . the results show that the current state-of-the-art chatbot can be easily goaded into making contradictions.

Copied to clipboard

Tianchi Bi, Hao Xiong, Zhongjun He, Hua Wu, Haifeng Wang

Autoregressive Pre-Training on Pixels and Texts (2024.emnlp-main)

Challenge:	Experimental results show that training with more than one agent improves translation quality and improves accuracy.
Approach:	They propose to introduce diverse agents in an in- teractive updating process to train NMT models with an additional agent.
Outcome:	The proposed approach improves on NIST Chinese-English, IWSLT 2014 German- English, WMT 2014 English-German translation tasks and shows competitive performance on all tasks.

Copied to clipboard

Yekun Chai, Qingyi Liu, Jingwu Xiao, Shuohuan Wang, Yu Sun, Hua Wu

Towards Conversational Recommendation over Multi-Type Dialogs (2020.acl-main)

Challenge:	pixel-based language modeling integrates visual and textual data to improve performance of language models.
Approach:	They propose a method that integrates visual and textual data into an autoregressive framework.
Outcome:	The proposed method improves performance of pixel-based language models by incorporating visual and textual data.

Copied to clipboard

Zeming Liu, Haifeng Wang, Zheng-Yu Niu, Hua Wu, Wanxiang Che, Ting Liu

Syntactic and Semantic-driven Learning for Open Information Extraction (2020.findings-emnlp)

Challenge:	In recent years, there has been a significant increase in the work of conversational recommendation due to the rise of voice-based bots.
Approach:	They use a Chinese dialog dataset DuRecDial to study conversational recommendation in the context of multi-type dialogs where bots can proactively lead a conversation from a non-recommendation dialog to a recommendation dialog.
Outcome:	The proposed dataset allows to investigate different parts of the overall problem, e.g., how to naturally lead a dialog, how interact with users for recommendation.

Copied to clipboard

Jialong Tang, Yaojie Lu, Hongyu Lin, Xianpei Han, Le Sun, Xinyan Xiao, Hua Wu

Reinforced Informativeness Optimization for Long-Form Retrieval-Augmented Generation (2026.acl-long)

Challenge:	Experimental results show that our approach significantly outperforms the supervised counterparts, and can even achieve competitive performance to supervised state-of-the-art (SoA) model.
Approach:	They propose a syntactic and semantic-driven learning approach that can learn open IE models without human-labelled data by leveraging syntakic and semantic knowledge as noisier, higher-level supervision.
Outcome:	The proposed approach outperforms supervised counterparts and can achieve competitive performance to supervised state-of-the-art models.

Copied to clipboard

Yuhao Wang, Ruiyang Ren, Yucheng Wang, Xin Zhao, Jing Liu, Hua Wu, Haifeng Wang

MoE Adapter for Large Audio Language Models: Sparsity, Disentanglement, and Gradient-Conflict-Free (2026.findings-acl)

Challenge:	Existing reinforcement learning systems lack verifiable reward mechanisms for long-form question answering . current systems lack reliable long-term answers due to lack of factual content .
Approach:	They propose a framework for reinforced verifiable informativeness optimization . it defines informativeness as measurable and externally verifier objective for RL .
Outcome:	Experiments show that RioRAG achieves higher factual recall and faithfulness . the proposed framework is based on a framework that uses nugget-centric verification with cross-source checks .

Copied to clipboard

Yishu Lei, Shuwei He, Hu Jing, Dan Zhang, Xianlong Luo, Danxiang Zhu, Shikun Feng, Rui Liu, Jingzhou HE, Yu Sun, Hua Wu, Haifeng Wang

Learning What Matters: Dynamic Dimension Selection and Aggregation for Interpretable Vision-Language Reward Modeling (2026.acl-long)

Challenge:	Existing research on Large Language Models (LLMs) limited to textual input modality . acoustic information is intrinsically heterogeneous, entangling attributes such as speech, music, and environmental context.
Approach:	They propose a sparse Mixture-of-Experts architecture to decouple acoustic information by routing audio tokens to specialized experts.
Outcome:	The proposed architecture outperforms existing models on audio semantic and paralinguistic tasks while retaining shared experts for global context.

Copied to clipboard

Qiyuan Chen, Hongsen Huang, Jiahe Chen, Qian Shao, Jintai Chen, Hongxia Xu, Renjie Hua, Ren Chuan, Jian Wu

PLANET: Dynamic Content Planning in Autoregressive Transformers for Long-form Text Generation (2022.acl-long)

Challenge:	Existing multimodal reward models are interpretable but slow, while discriminative ones are opaque "black boxes."
Approach:	They propose a framework that dynamically decomposes evaluation into granular, interpretable dimensions.
Outcome:	The proposed framework outperforms open-source reward models on benchmarks like VL-RewardBench.

Copied to clipboard

Zhe Hu, Hou Pong Chan, Jiachen Liu, Xinyan Xiao, Hua Wu, Lifu Huang

SgSum:Transforming Multi-document Summarization into Sub-graph Selection (2021.emnlp-main)

Challenge:	Existing methods for text generation still suffer from incoherence problems . Neural sequence-to-sequence (seq2sequ) models generate fluent results .
Approach:	They propose a novel generation framework that leverages autoregressive self-attention mechanism to conduct content planning and surface realization dynamically.
Outcome:	The proposed framework outperforms baseline models and generates more coherent texts with richer contents.

Copied to clipboard

Moye Chen, Wei Li, Jiachen Liu, Xinyan Xiao, Hua Wu, Haifeng Wang

Investigating the Factual Knowledge Boundary of Large Language Models with Retrieval Augmentation (2025.coling-main)

Challenge:	Existing extractive multi-document summarization methods score each sentence individually and extract salient sentences one by one.
Approach:	They propose a novel framework for extractive multi-document summarization that selects a sub-graph as the summary instead of selecting salient sentences.
Outcome:	The proposed framework improves on existing methods on multi-document datasets and human evaluations show it produces more coherent and informative summaries.

Copied to clipboard

Ruiyang Ren, Yuhao Wang, Yingqi Qu, Wayne Xin Zhao, Jing Liu, Hua Wu, Ji-Rong Wen, Haifeng Wang

Towards Boosting the Open-Domain Chatbot with Human Feedback (2023.acl-long)

Challenge:	Large language models (LLMs) have shown impressive prowess in solving a wide range of tasks with world knowledge, but it remains unclear how well they perceive their factual knowledge boundaries.
Approach:	They propose to use a retrieval augmentation approach to enhance LLMs' awareness of factual knowledge boundaries to analyze factual and factual information in open-domain question answering (QA)
Outcome:	The proposed method improves LLMs’ QA and judgemental capabilities by integrating supporting documents with the questions.

Copied to clipboard

Hua Lu, Siqi Bao, Huang He, Fan Wang, Hua Wu, Haifeng Wang

A Multitask Active Learning Framework for Natural Language Understanding (2020.coling-main)

Challenge:	Existing frameworks for pre-training open-domain dialogue models with social media comments generate coherent replies but have difficulties producing engaging responses.
Approach:	They propose a framework to boost the open-domain chatbot by leveraging human feedback and annotating the model's candidate responses.
Outcome:	The proposed framework boosts the open-domain chatbot by leveraging human demonstrated responses and leveraging the implicit preference in the data collection process.

Copied to clipboard

Hua Zhu, Wu Ye, Sihan Luo, Xidong Zhang

RocketQAv2: A Joint Training Method for Dense Passage Retrieval and Passage Re-ranking (2021.emnlp-main)

Challenge:	Existing studies have shown that relation information between intents and slots can improve the efficiency of active learning algorithms.
Approach:	They propose a multitask active learning framework that exploits relation information between sub-tasks provided by a joint model.
Outcome:	The proposed framework achieves competitive performance with less training data than baseline methods on all datasets.

Copied to clipboard

Ruiyang Ren, Yingqi Qu, Jing Liu, Wayne Xin Zhao, QiaoQiao She, Hua Wu, Haifeng Wang, Ji-Rong Wen

HFT: Half Fine-Tuning for Large Language Models (2025.acl-long)

Challenge:	Recent studies show that passage retrieval and passage reranking are important for achieving mutual improvement.
Approach:	They propose a unified listwise training approach for passage retrieval and passage reranking that incorporates a retrieval procedure and a hybrid data augmentation strategy.
Outcome:	The proposed approach improves on both MSMARCO and Natural Questions datasets.

Copied to clipboard

Tingfeng Hui, Zhenyu Zhang, Shuohuan Wang, Weiran Xu, Yu Sun, Hua Wu

CoTKR: Chain-of-Thought Enhanced Knowledge Rewriting for Complex Knowledge Graph Question Answering (2024.emnlp-main)

Challenge:	Large language models (LLMs) with one or more fine-tuning phases can unlock various capabilities, but can be catastrophic forgetting during sequential training.
Approach:	They propose a method to regularly reset partial parameters to mitigate forgetting issues by using half fine-tuning instead of full fine-uning.
Outcome:	The proposed approach reduces the risk of catastrophic forgetting during training and the parametric knowledge lost during training may be overwhelmed by incoming training data.

Copied to clipboard

Yike Wu, Yi Huang, Nan Hu, Yuncheng Hua, Guilin Qi, Jiaoyan Chen, Jeff Pan

HAIC: Improving Human Action Understanding and Generation with Better Captions for Multi-modal Large Language Models (2025.acl-long)

Challenge:	Existing knowledge rewriting methods may include irrelevant information, omit crucial details, or fail to align with the question’s semantics.
Approach:	They propose a new rewriting method CoTKR for generating reasoning traces and corresponding knowledge in an interleaved manner, thereby mitigating the limitations of single-step knowledge rewrite.
Outcome:	The proposed method mitigates the limitations of single-step knowledge rewriting and bridges the preference gap between the knowledge reactor and the question answering (QA) model.

Copied to clipboard

Xiao Wang, Jingyun Hua, Weihong Lin, Yuanxing Zhang, Fuzheng Zhang, Jianlong Wu, Di Zhang, Liqiang Nie

Knowledge Aware Conversation Generation with Explainable Reasoning over Augmented Graphs (D19-1)

Challenge:	Existing studies have shown that high-quality video captions can improve MLLMs' performance on videos involving human actions.
Approach:	They propose a data annotation pipeline to collect videos featuring clear human actions from the Internet and annotate them in a standardized caption format that uses human attributes to distinguish individuals.
Outcome:	The proposed pipeline combines two datasets to evaluate human action understanding.

Copied to clipboard

Zhibin Liu, Zheng-Yu Niu, Hua Wu, Haifeng Wang

On Training Data Influence of GPT Models (2024.emnlp-main)

Challenge:	Existing knowledge-based open domain conversation generation models are limited by the use of unstructured knowledge texts.
Approach:	They propose a knowledge aware chatting machine with three components, an augmented knowledge graph with both triples and texts, knowledge selector, and knowledge aware response generator.
Outcome:	The proposed system is more explainable and flexible than state-of-the-art models.

Copied to clipboard

Yekun Chai, Qingyi Liu, Shuohuan Wang, Yu Sun, Qiwei Peng, Hua Wu

CH-SIMS: A Chinese Multimodal Sentiment Analysis Dataset with Fine-grained Annotation of Modality (2020.acl-main)

Challenge:	generative language models have redefined performance standards across tasks . current research on the influence of training data on autoregressivity remains underexplored .
Approach:	They propose a parameterized simulation to assess the impact of training examples on the training dynamics of GPT models.
Outcome:	The proposed approach compares existing methods with existing methods across training scenarios in generative language models, spanning tasks across 14 million to 2.8 billion parameters.

Copied to clipboard

Wenmeng Yu, Hua Xu, Fanyang Meng, Yilin Zhu, Yixiao Ma, Jiele Wu, Jiyun Zou, Kaicheng Yang

Are Shortest Rationales the Best Explanations for Human Understanding? (2022.acl-short)

Challenge:	Existing studies in multimodal sentiment analysis only use unified multimodal annotations, which do not reflect the independent sentiment of single modalities.
Approach:	They propose a Chinese single- and multi-modal sentiment analysis dataset with multimodal and independent unimodal annotations that can be used to study the interaction between modalities.
Outcome:	The proposed methods achieve state-of-the-art performance and learn more distinctive unimodal representations.

Copied to clipboard

Hua Shen, Tongshuang Wu, Wenbo Guo, Ting-Hao Huang

Bi-SimCut: A Simple Strategy for Boosting Neural Machine Translation (2022.naacl-main)

Challenge:	Existing models favor extracting the shortest possible rationales to explain model predictions . however, this assumption has yet to be validated .
Approach:	They propose a model that extracts rationales at any target length from text inputs . they show that rationale lengths too short do not help humans predict labels better .
Outcome:	The proposed model achieves compatible end-task performance and human-annotated rationale agreement compared to baseline models .

Copied to clipboard

Pengzhi Gao, Zhongjun He, Hua Wu, Haifeng Wang

Long Time No See! Open-Domain Conversation with Long-Term Persona Memory (2022.findings-acl)

Challenge:	BLEU scores of 31.16 for ende and 38.37 for deen on the IWSLT14 dataset, 30.78 for entde, 35.15 for de en and 27.17 for zhen .
Approach:	They propose a bidirectional pretraining and unidirectional finetuning procedure to boost NMT performance.
Outcome:	The proposed method achieves strong translation performance across five datasets.

Copied to clipboard

Xinchao Xu, Zhibin Gou, Wenquan Wu, Zheng-Yu Niu, Hua Wu, Haifeng Wang, Shihang Wang

AlignX: Advancing Multilingual Large Language Models with Multilingual Representation Alignment (2025.emnlp-main)

Challenge:	Existing persona dialogue datasets and models can build long-term relationships with humans . however, current open-domain dialogue systems cannot build long relationships with users .
Approach:	They propose a long-term memory conversation dataset and a dialogue generation framework with long-Term memory mechanism to extract and continuously update long-time persona memory.
Outcome:	The proposed system outperforms baselines in terms of long-term dialogue consistency . the proposed system can build long-lasting relationships between humans and bots .

Copied to clipboard

Mengyu Bu, Shaolei Zhang, Zhongjun He, Hua Wu, Yang Feng

Multi-Passage Machine Reading Comprehension with Cross-Passage Answer Verification (P18-1)

Challenge:	Multilingual large language models (LLMs) possess impressive multilingual understanding and generation capabilities, but performance and cross-lingual alignment often lag for non-dominant languages.
Approach:	They propose a representation-level framework to enhance multilingual performance of pre-trained LLMs by integrating multilingual semantic alignment and language feature integration.
Outcome:	The proposed framework improves multilingual capability of pre-trained LLMs by bringing representations closer and improving cross-lingual alignment.

Copied to clipboard

Yizhong Wang, Kai Liu, Jing Liu, Wei He, Yajuan Lyu, Hua Wu, Sujian Li, Haifeng Wang

Correcting Chinese Spelling Errors with Phonetic Pre-training (2021.findings-acl)

Challenge:	Recent years have seen rapid growth in the MRC community . MRC is believed to be a crucial step in building a general intelligent agent .
Approach:	They propose an end-to-end neural model that enables multiple passages to verify each other based on their content representations.
Outcome:	The proposed model outperforms the baseline on the English MS-MARCO dataset and the Chinese DuReader dataset, and achieves state-of-the-art performance on both datasets.

Copied to clipboard

Ruiqing Zhang, Chao Pang, Chuanqiang Zhang, Shuohuan Wang, Zhongjun He, Yu Sun, Hua Wu, Haifeng Wang

Syntactic Text Processing

Challenge:	Existing methods for Chinese spelling correction only use pre-trained language model or incorporate phonological information as external knowledge.
Approach:	They propose a phonetic Chinese spelling correction model that integrates phonetic features into language model by leveraging pre-training and fine-tuning methods.
Outcome:	The proposed model outperforms existing methods on SIGHAN datasets and improves on other datasets.

Few-Shot Complex Knowledge Base Question Answering via Meta Reinforcement Learning (2020.emnlp-main)

Copied to clipboard

Yuncheng Hua, Yuan-Fang Li, Gholamreza Haffari, Guilin Qi, Tongtong Wu

DuRecDial 2.0: A Bilingual Parallel Corpus for Conversational Recommendation (2021.emnlp-main)

Challenge:	Existing approaches to complex question-answering (CQA) exhibit uneven performance when questions have different types, harboring inherently different characteristics, e.g., difficulty level.
Approach:	They propose a meta-reinforcement learning approach to program induction in CQA to tackle the potential distributional bias in questions.
Outcome:	The proposed method achieves state-of-the-art performance on the CQA dataset while using only five trial trajectories for the top-5 retrieved questions in each support set.

Copied to clipboard

Zeming Liu, Haifeng Wang, Zheng-Yu Niu, Hua Wu, Wanxiang Che

D-NET: A Pre-Training and Fine-Tuning Framework for Improving the Generalization of Machine Reading Comprehension (D19-58)

Challenge:	Existing datasets for conversational recommendation are limited to English and Chinese .
Approach:	They propose a bilingual parallel human-to-human recommendation dialog dataset . the data item is annotated in two languages, both English and Chinese .
Outcome:	The proposed dataset provides a testbed for future studies of multilingual and cross-lingual conversational recommendation.

Copied to clipboard

Hongyu Li, Xiyuan Zhang, Yibing Liu, Yiming Zhang, Quan Wang, Xiangyang Zhou, Jing Liu, Hua Wu, Haifeng Wang

Learning Multilingual Sentence Representations with Cross-lingual Consistency Regularization (2023.emnlp-industry)

Challenge:	MRC requires machines to understand text and answer questions about the text.
Approach:	They propose a simple system Baidu submitted for MRQA 2019 Shared Task that focused on generalization of machine reading comprehension (MRC) models.
Outcome:	The proposed system is ranked at top 1 of all participants in terms of averaged F1 score.

Copied to clipboard

Pengzhi Gao, Liwen Zhang, Zhongjun He, Hua Wu, Haifeng Wang

UNIMO-2: End-to-End Unified Vision-Language Grounded Learning (2022.findings-acl)

Challenge:	Experimental results on multilingual similarity search and bitext mining tasks show the effectiveness of our approach.
Approach:	They propose a multilingual sentence representation model that aligns different languages in a shared representation space.
Outcome:	The proposed model performs better than LASER3 on similarity searches and bitext mining tasks.

Copied to clipboard

Wei Li, Can Gao, Guocheng Niu, Xinyan Xiao, Hao Liu, Jiachen Liu, Hua Wu, Haifeng Wang

A Thorough Examination on Zero-shot Dense Retrieval (2023.findings-emnlp)

Challenge:	Existing methods for vision-language pre-training can only learn from aligned image-caption data and rely heavily on expensive regional features.
Approach:	They propose an end-to-end unified-modal pre-training framework for joint learning . they propose to conduct grounded learning on both images and texts via a sharing grounded space .
Outcome:	The proposed model improves visual and visual semantic alignment on images and texts.

Copied to clipboard

Ruiyang Ren, Yingqi Qu, Jing Liu, Xin Zhao, Qifei Wu, Yuchen Ding, Hua Wu, Haifeng Wang, Ji-Rong Wen

Unleashing Spatial Reasoning in Multimodal Large Language Models via Textual Representation Guided Reasoning (2026.acl-long)

Challenge:	Recent advances in dense retrieval (DR) models have been shown to be not as competitive as traditional sparse retrieval models in a zero-shot retrieval setting.
Approach:	They propose to examine the zero-shot capability of DR models by analyzing key factors related to source training set and potential bias from target dataset.
Outcome:	The proposed model is not as competitive as sparse retrieval models in a zero-shot retrieval setting.

Copied to clipboard

Jiacheng Hua, Yishu Yin, Yuhang Wu, Tai Wang, Yifei Huang, Miao Liu

RuleArena: A Benchmark for Rule-Guided Reasoning with LLMs in Real-World Scenarios (2025.acl-long)

Challenge:	Existing Multimodal Large Language Models struggle with 3D spatial reasoning as they fail to construct structured abstractions of the 3D environment depicted in video inputs.
Approach:	They propose a prompting method that induces MLLMs to generate 3D representations as reasoning traces for more accurate spatial question answering.
Outcome:	Extensive experiments on VSI-Bench and OST-Bech show that TRACE improves over prior prompting strategies across a diverse range of MLLM backbones.

Copied to clipboard

Ruiwen Zhou, Wenyue Hua, Liangming Pan, Sitao Cheng, Xiaobao Wu, En Yu, William Yang Wang

Mixup Decoding for Diverse Machine Translation (2021.findings-emnlp)

Challenge:	RuleArena assesses the ability of large language models (LLMs) to follow complex, real-world rules in reasoning.
Approach:	They propose a benchmark to evaluate the ability of large language models (LLMs) to follow complex, real-world rules in reasoning.
Outcome:	The proposed benchmark covers airline baggage fees, NBA transactions, and tax regulations.

Copied to clipboard

Jicheng Li, Pengzhi Gao, Xuanfu Wu, Yang Feng, Zhongjun He, Hua Wu, Haifeng Wang

Machine Translation

Challenge:	Existing methods for generating multiple translations for source and target languages neglect the one-to-many mapping between the source and the target languages.
Approach:	They propose a method to generate different translations for the input sentence by linearly interpolating it with different sentence pairs sampled from the training corpus during decoding.
Outcome:	Experiments on WMT’16 en-ro, WMT'14 en de, and WMT ‘17 zh-en show that the proposed method outperforms all previous diverse machine translation methods.

Improving Zero-shot Multilingual Neural Machine Translation by Leveraging Cross-lingual Consistency Regularization (2023.findings-acl)

Copied to clipboard

Pengzhi Gao, Liwen Zhang, Zhongjun He, Hua Wu, Haifeng Wang

Fine-grained Entity Typing via Label Reasoning (2021.emnlp-main)

Challenge:	Existing methods to improve zero-shot translation performance by learning language-agnostic representations and maximizing cross-lingual transfer have been proposed.
Approach:	They propose a cross-lingual consistency regularization to bridge the representation gap between different languages and boost zero-shot translation performance.
Outcome:	The proposed model improves translation performance on low-resource and high-res benchmarks and closes the sentence representation gap and aligns the representation space.

Copied to clipboard

Qing Liu, Hongyu Lin, Xinyan Xiao, Xianpei Han, Le Sun, Hua Wu

DuQM: A Chinese Dataset of Linguistically Perturbed Natural Questions for Evaluating the Robustness of Question Matching Models (2022.emnlp-main)

Challenge:	Existing approaches to fine-grained entity typing are based on independent classification paradigms, which make them difficult to recognize inter-dependent, long-tailed and fine-granular entities.
Approach:	They propose a label reasoning network that exploits label dependencies knowledge entailed in the data.
Outcome:	The proposed network can model, learn and reason complex labels in a sequence-to-set, end-to end manner.

Copied to clipboard

Hongyu Zhu, Yan Chen, Jing Yan, Jing Liu, Yu Hong, Ying Chen, Hua Wu, Haifeng Wang

Robustness in NLP

Challenge:	a comprehensive evaluation of QM models should be conducted on natural texts, not on artificial adversarial examples . ral models are often not robust to adversarials, which means they predict unexpected outputs .
Approach:	They use a Chinese dataset to evaluate the robustness of QM models . they show that the effect of artificial adversarial examples does not work on natural texts .
Outcome:	The proposed model is more robust than other models on natural questions with 32 linguistic perturbations.

Icon2: Aligning Large Language Models Using Self-Synthetic Preference Data via Inherent Regulation (2025.emnlp-main)

Copied to clipboard

Qiyuan Chen, Hongsen Huang, Qian Shao, Jiahe Chen, Jintai Chen, Hongxia Xu, Renjie Hua, Ren Chuan, Jian Wu

BASS: Boosting Abstractive Summarization with Unified Semantic Graph (2021.acl-long)

Challenge:	Large Language Models (LLMs) require high quality preference datasets to align with human preferences.
Approach:	They propose a framework that leverages inherent regulation of LLMs’ representation space for efficient and tailored preference dataset construction, named Icon2.
Outcome:	The proposed framework improves performance on benchmarks like AlpacaEval 2.0 and Arena-Hard while reducing computational costs by up to 48.1%.

Copied to clipboard

Wenhao Wu, Wei Li, Xinyan Xiao, Jiachen Liu, Ziqiang Cao, Sujian Li, Hua Wu, Haifeng Wang

DuSQL: A Large-Scale and Pragmatic Chinese Text-to-SQL Dataset (2020.emnlp-main)

Challenge:	Abstractive summarization for long-document or multi-document remains challenging for Seq2Seq as it does not analyze long-distance relations in text.
Approach:	They propose a framework for Boosting Abstractive Summarization based on a unified Semantic graph which aggregates co-referent phrases distributing across a long range of context and conveys rich relations between phrases.
Outcome:	The proposed framework improves document representation and summary generation process by leveraging the graph structure.

Copied to clipboard

Lijie Wang, Ao Zhang, Kun Wu, Ke Sun, Zhenghua Li, Hua Wu, Min Zhang, Haifeng Wang

Programming Languages in NLP

Challenge:	Existing text-to-SQL parsing methods mainly focus on English, but there is no labeled data available for the language . a larges-scale and pragmatic Chinese dataset is used for cross-domain text- to-Sql task .
Approach:	They propose a larges-scale Chinese dataset for a cross-domain text-to-SQL task . they analyze questions from several representative applications and use an effective data construction framework .
Outcome:	The proposed dataset contains 200 databases, 813 tables, and 23,797 question/SQL pairs.

Weights-Rotated Preference Optimization for Large Language Models (2025.emnlp-main)

Copied to clipboard

Chenxu Yang, Ruipeng Jia, Mingyu Zheng, Naibin Gu, Zheng Lin, Siyuan Chen, Weichong Yin, Hua Wu, Weiping Wang

Data Augmentation with Hierarchical SQL-to-Question Generation for Cross-domain Text-to-SQL Parsing (2021.emnlp-main)

Challenge:	Existing methods to align large language models with high reward hacking are limited by the complexity of the parameter space and the complexity.
Approach:	They propose a weights-rotated preference optimization algorithm that constrains the output layer logits with the KL divergence inherited from DPO and fine-tunes the intermediate hidden states.
Outcome:	The proposed algorithm achieves a 3.27-point improvement on AlpacaEval 2 and surpasses the best baseline by 6.2 to 7.5 points on MT-Bench with merely 0.015% of the trainable parameters.

Copied to clipboard

Kun Wu, Lijie Wang, Zhenghua Li, Ao Zhang, Xinyan Xiao, Hua Wu, Min Zhang, Haifeng Wang

Programming Languages in NLP

Challenge:	Existing approaches to parse text-to-SQL data are lacking labeled data for unseen evaluation databases.
Approach:	They propose a framework for enhancing SQL queries by automatically producing large numbers of SQL queries based on an abstract syntax tree grammar.
Outcome:	The proposed framework can produce high-quality natural language questions over strong baselines.

An Empirical Study of Consistency Regularization for End-to-End Speech-to-Text Translation (2024.naacl-long)

Copied to clipboard

Pengzhi Gao, Ruiqing Zhang, Zhongjun He, Hua Wu, Haifeng Wang

Text Generation

Challenge:	Existing methods for speech-to-text translation (ST) have achieved impressive supervised and zero-shot performance.
Approach:	They propose to use consistency regularization methods to boost end-to-end (E2E) speech-totext translation (ST) by regularizing the intra-modal consistency instead of the modality gap.
Outcome:	The proposed training strategies achieve state-of-the-art (SOTA) performance in most translation directions.

ERNIE-Code: Beyond English-Centric Cross-lingual Pretraining for Programming Languages (2023.findings-acl)

Copied to clipboard

Yekun Chai, Shuohuan Wang, Chao Pang, Yu Sun, Hao Tian, Hua Wu

BASES: Large-scale Web Search User Simulation with Large Language Model based Agents (2024.findings-emnlp)

Challenge:	ERNIE-Code is a unified pre-trained language model for 116 NLs and 6 PLs.
Approach:	They propose a unified pre-trained language model for 116 NLs and 6 PLs . they employ span-corruption language modeling that learns patterns from monolingual NL or PL .
Outcome:	The proposed model outperforms previous multilingual models for NL or NL across end tasks.

Copied to clipboard

Ruiyang Ren, Peng Qiu, Yingqi Qu, Jing Liu, Xin Zhao, Hua Wu, Ji-Rong Wen, Haifeng Wang

Sparse Growing Transformer: Training-Time Sparse Depth Allocation via Progressive Attention Looping (2026.findings-acl)

Challenge:	Existing research on web search rely on real-user experiments, which can be costly to scale up.
Approach:	They propose a user simulation framework with LLM-based agents that can generate unique user profiles at scale.
Outcome:	The proposed framework can generate unique user profiles at scale, leading to diverse search behaviors.

Copied to clipboard

Yao Chen, Yilong Chen, Yinqi Yang, Junyuan Shang, Zhenyu Zhang, Zefeng Zhang, Shuaiyi Nie, Shuohuan Wang, Yu Sun, Hua Wu, Haifeng Wang, Tingwen Liu

DU-VLG: Unifying Vision-and-Language Generation via Dual Sequence-to-Sequence Pre-training (2022.findings-acl)

Challenge:	Existing approaches to increasing effective depth of LLMs rely on parameter reuse, extending computation through recursive execution.
Approach:	They propose a training-time sparse depth allocation framework that progressively increases depth for a small subset of parameters as training evolves.
Outcome:	The proposed model outperforms existing approaches to increasing the effective depth of language models while reducing training FLOPs overhead from approximately 16–20% to only 1–3% relative to a standard Transformer backbone.

Copied to clipboard

Luyang Huang, Guocheng Niu, Jiachen Liu, Xinyan Xiao, Hua Wu

Watermarking PLMs on Classification Tasks by Combining Contrastive Learning with Weight Perturbation (2023.findings-emnlp)

Challenge:	Existing vision-and-language generation models cannot utilize pair-wise images and text through bi-directional generation due to the limitations of the model structure and pre-training objectives.
Approach:	They propose a framework which unifies vision-and-language generation as sequence generation problems.
Outcome:	The proposed framework achieves better performance than variants trained with uni-directional generation objectives or the variant without the commitment loss on image captioning and text-to-image generation datasets.

Copied to clipboard

Chenxi Gu, Xiaoqing Zheng, Jianhan Xu, Muling Wu, Cenyuan Zhang, Chengsong Huang, Hua Cai, Xuanjing Huang

BeamLoRA: Beam-Constraint Low-Rank Adaptation (2025.acl-long)

Challenge:	Large pre-trained language models (PLMs) are highly valuable intellectual property due to their expensive training costs.
Approach:	They propose to embed backdoors that can be triggered by specific inputs into models by model watermarking.
Outcome:	The proposed method can be used to protect the intellectual property of large pre-trained language models without knowledge about downstream tasks.

Copied to clipboard

Naibin Gu, Zhenyu Zhang, Xiyu Liu, Peng Fu, Zheng Lin, Shuohuan Wang, Yu Sun, Hua Wu, Weiping Wang, Haifeng Wang

ERNIE-Gram: Pre-Training with Explicitly N-Gram Masked Language Modeling for Natural Language Understanding (2021.naacl-main)

Challenge:	Low-Rank Adaptation (LoRA) is one of the most efficient parameter-efficient fine-tuning methods.
Approach:	They propose to conceptualize each LoRA module as a beam where each rank corresponds to a potential sub-solution.
Outcome:	The proposed method improves performance on three base models and 12 datasets.

Copied to clipboard

Dongling Xiao, Yu-Kun Li, Han Zhang, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang

ARNOR: Attention Regularization based Noise Reduction for Distant Supervision Relation Classification (P19-1)

Challenge:	Existing methods to model coarse-grained linguistic information do not integrate coarse-gram information into pre-training.
Approach:	They propose an explicitly n-gram masking method to enhance integration of coarse-grained linguistic information into pre-training.
Outcome:	The proposed method outperforms existing models on English and Chinese text corpora and fine-tunes on 19 downstream tasks.

Copied to clipboard

Wei Jia, Dai Dai, Xinyan Xiao, Hua Wu

Explainability & Interpretability in NLP

Challenge:	Distant supervision is used for relation classification but it introduces noisy labels . a novel approach to distant supervision relation classification is proposed .
Approach:	They propose a framework for distant supervision relation classification using attention regularization and attention regularizing . they assume that a trustable relation label should be explained by the neural attention model .
Outcome:	The proposed framework improves on the NYT data and noise reduction effect over state-of-the-art methods.

Non-Autoregressive Chinese ASR Error Correction with Phonological Training (2022.naacl-main)

Copied to clipboard

Zheng Fang, Ruiqing Zhang, Zhongjun He, Hua Wu, Yanan Cao

Speech & Audio in NLP

Challenge:	Existing methods to correct ASR errors focus on fixed-length corrections, but rarely consider variable-length ones.
Approach:	They propose a non-autoregressive method to correct Chinese ASR errors . they use phonological tokens to extend the source sentence for variable-length correction .
Outcome:	The proposed method improves word error rate and speeds up inference by 6.2 times compared with the autoregressive model.

IAEval: A Comprehensive Evaluation of Instance Attribution on Natural Language Understanding (2023.findings-emnlp)

Copied to clipboard

Peijian Gu, Yaozong Shen, Lijie Wang, Quan Wang, Hua Wu, Zhendong Mao

Challenge:	Instance attribution (IA) aims to identify the training instances leading to the prediction of a test example.
Approach:	They propose a systematic and comprehensive evaluation scheme covering four significant requirements: sufficiency, completeness, stability and plausibility.
Outcome:	The proposed evaluation scheme covers four significant requirements: sufficiency, completeness, stability and plausibility.

PLATO: Pre-trained Dialogue Generation Model with Discrete Latent Variable (2020.acl-main)

Copied to clipboard

Siqi Bao, Huang He, Fan Wang, Hua Wu, Haifeng Wang

Golden Touchstone: A Comprehensive Bilingual Benchmark for Evaluating Financial Large Language Models (2025.findings-emnlp)

Challenge:	Existing pre-training models for dialogue generation have been proven effective for a wide range of tasks.
Approach:	They propose a dialogue generation pre-training framework that leverages bi-directional context and uni-directional characteristic of language generation.
Outcome:	The proposed framework is superior to existing models on three publicly available datasets.

Copied to clipboard

Xiaojun Wu, Junxi Liu, Huan-Yi Su, Zhouchi Lin, Yiyan Qi, Chengjin Xu, Jiajun Su, Jiajie Zhong, Fuwei Wang, Saizhuo Wang, Fengrui Hua, Jia Li, Jian Guo

Curiosity-Driven Reinforcement Learning from Human Feedback (2025.acl-long)

Challenge:	Existing financial benchmarks suffer from limited language and task coverage, low-quality datasets, and inadequate adaptability for LLM evaluation.
Approach:	They propose a bilingual benchmark for financial LLMs that assesses models’ language understanding and generation capabilities.
Outcome:	The proposed bilingual benchmark assesses models’ language understanding and generation capabilities.

Copied to clipboard

Haoran Sun, Yekun Chai, Shuohuan Wang, Yu Sun, Hua Wu, Haifeng Wang

Challenge:	Reinforcement learning from human feedback (RLHF) has proven effective in aligning large language models with human preferences, but often at the cost of reduced output diversity.
Approach:	They propose a framework that incorporates intrinsic rewards for novel states alongside traditional sparse extrinsic rewards to optimize both output diversity and alignment quality.
Outcome:	The proposed framework achieves significant gains in diversity on multiple diversity-oriented metrics while maintaining alignment with human preferences comparable to standard RLHF.

IBADR: an Iterative Bias-Aware Dataset Refinement Framework for Debiasing NLU models (2023.emnlp-main)

Copied to clipboard

Xiaoyue Wang, Xin Liu, Lijie Wang, Yaoxiang Wang, Jinsong Su, Hua Wu

Upcycling Instruction Tuning from Dense to Mixture-of-Experts via Parameter Merging (2025.acl-long)

Challenge:	Using manual data analysis, dataset refinement approaches are often unable to cover all the potential biased features.
Approach:	They propose an iterative bias-aware dataset refinement framework which debiases NLU models without predefining biased features.
Outcome:	The proposed framework outperforms existing methods and is compatible with model-centric methods.

Copied to clipboard

Tingfeng Hui, Zhenyu Zhang, Shuohuan Wang, Yu Sun, Hua Wu, Sen Su

Know More about Each Other: Evolving Dialogue Strategy via Compound Assessment (P19-1)

Challenge:	Existing methods for tuning large language models from dense to MoE face significant data requirements and require large-scale post-training.
Approach:	They propose an upcycling instruction tuning approach for tuning a dense pre-trained model into a MoE instruction model using genetic algorithm and parameter merging.
Outcome:	The proposed approach improves the performance of large language models with a small amount of seed data and improves their scaling.

Copied to clipboard

Siqi Bao, Huang He, Fan Wang, Rongzhong Lian, Hua Wu

Enhancing Pre-Trained Language Representations with Rich Knowledge for Machine Reading Comprehension (P19-1)

Challenge:	Existing approaches to generate informative responses based on external knowledge are limited to singleround settings.
Approach:	They propose a framework for multi-turn conversations with two dialogue agents . they propose to evaluate dialogues on informativeness and coherence .
Outcome:	The proposed framework outperforms state-of-the-art approaches significantly on the publicly available dataset.

Copied to clipboard

An Yang, Quan Wang, Jing Liu, Kai Liu, Yajuan Lyu, Hua Wu, Qiaoqiao She, Sujian Li

ERNIE-Layout: Layout Knowledge Enhanced Pre-training for Visually-rich Document Understanding (2022.findings-emnlp)

Challenge:	Recent results show pre-trained language models (LMs) can improve machine reading comprehension (MRC) Experimental results indicate that KT-NET offers significant and consistent improvements over BERT .
Approach:	They propose a method that leverages external knowledge bases to improve machine reading comprehension (MRC) KT-NET employs an attention mechanism to select desired knowledge from KBs and fuses selected knowledge with BERT to enable context- and knowledge-aware predictions.
Outcome:	The proposed model outperforms baseline models on ReCoRD and SQuAD1.1 benchmarks and ranks 1st on the ReCoDR and SQUAD1.1 leaderboards.

Copied to clipboard

Qiming Peng, Yinxu Pan, Wenjin Wang, Bin Luo, Zhenyu Zhang, Zhengjie Huang, Yuhui Cao, Weichong Yin, Yongfeng Chen, Yin Zhang, Shikun Feng, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang

RocketQA: An Optimized Training Approach to Dense Passage Retrieval for Open-Domain Question Answering (2021.naacl-main)

Challenge:	Existing methods for visually rich document understanding lack layout-centered knowledge . experimental results show that ERNIE-Layout improves layout awareness .
Approach:	They propose a document pre-training solution with layout knowledge enhancement in the whole workflow to learn better representations that combine the features from text, layout, and image.
Outcome:	The proposed model outperforms existing models on key downstream tasks.

Copied to clipboard

Yingqi Qu, Yuchen Ding, Jing Liu, Kai Liu, Ruiyang Ren, Wayne Xin Zhao, Daxiang Dong, Hua Wu, Haifeng Wang

Query Enhanced Knowledge-Intensive Conversation via Unsupervised Joint Modeling (2023.acl-long)

Challenge:	Open-domain question answering uses dense passage retrieval to find answers . however, it is difficult to effectively train a dual-encoder due to discrepancy between training and inference .
Approach:	They propose an optimized training approach to improve dense passage retrieval using RocketQA . they propose cross-batch negatives, denoised hard negatives and data augmentation .
Outcome:	The proposed approach outperforms state-of-the-art models on both MSMARCO and Natural Questions.

Copied to clipboard

Mingzhu Cai, Siqi Bao, Xin Tian, Huang He, Fan Wang, Hua Wu

PLATO-2: Towards Building an Open-Domain Chatbot via Curriculum Learning (2021.findings-acl)

Challenge:	Existing methods to retrieve knowledge-intensive conversations are based on external resources such as Wikipedia databases or search engine results.
Approach:	They propose an unsupervised query enhanced approach for knowledge-intensive conversations . they conduct experiments on three knowledge- intensive conversation datasets .
Outcome:	The proposed approach performs better than all unsupervised methods across three datasets and achieves competitive performance compared to supervised methods.

Copied to clipboard

Siqi Bao, Huang He, Fan Wang, Hua Wu, Haifeng Wang, Wenquan Wu, Zhen Guo, Zhibin Liu, Xinchao Xu

Exploring Contextual Word-level Style Relevance for Unsupervised Style Transfer (2020.acl-main)

Challenge:	PLATO-2 is a high-quality open-domain chatbot that can generate one-to-many mappings and improve response quality.
Approach:	They propose a curriculum learning process to build a high-quality open-domain chatbot . they use a coarse-grained generation model and latent variables to train a generative model .
Outcome:	The proposed model improves on Chinese and English data and can generate diverse responses and select the best response.

Copied to clipboard

Chulun Zhou, Liangyu Chen, Jiachen Liu, Xinyan Xiao, Jinsong Su, Sheng Guo, Hua Wu

Low-Resource NLP

Challenge:	Existing methods to unsupervised style transfer lack fine-grained control of the influence from the target style.
Approach:	They propose a model that exploits the relevance of each output word to the target style . they pretrain a style classifier and train an attentional Seq2seq model to reconstruct input sentences .
Outcome:	The proposed model achieves state-of-the-art performance in terms of transfer accuracy and content preservation.

Continual Machine Reading Comprehension via Uncertainty-aware Fixed Memory and Adversarial Domain Adaptation (2022.findings-naacl)

Copied to clipboard

Zhijing Wu, Hua Xu, Jingliang Fang, Kai Gao

Towards Zero-Shot Persona Dialogue Generation with In-Context Learning (2023.findings-acl)

Challenge:	Existing stationary-trained MRC systems are usually trained with in-domain data but are applied to new domain data.
Approach:	They propose a continual machine reading comprehension model with uncertainty-aware fixed memory and adversarial domain adaptation that keeps a stable understanding by learning both memory and new domain data.
Outcome:	The proposed model is superior to strong baselines and has a substantial incremental learning ability without catastrophically forgetting under two different continual MRC settings.

Copied to clipboard

Xinchao Xu, Zeyang Lei, Wenquan Wu, Zheng-Yu Niu, Hua Wu, Haifeng Wang