Papers by Hu Junjie

Single Sequence Prediction over Reasoning Graphs for Multi-hop QA (2023.acl-long)

Challenge:	Existing approaches to neural machine translation (NMT) are dependent on limited parallel data, and can be difficult to use for many language pairs.
Approach:	They propose a method where target-language sentences are re-ordered to match the order of the source and used as an additional source of training-time supervision.
Outcome:	The proposed method improves on simulated low-resource Japanese-to-English and real low-demand Uyghur-to English scenarios.

Copied to clipboard

Gowtham Ramesh, Makesh Narsimhan Sreedhar, Junjie Hu

Structured Data in NLP

Challenge:	Recent generative approaches for multi-hop question answering (QA) use fusion-in-decoder to generate a single sequence output . but, they often have difficulty accurately identifying passages corresponding to key entities in the context .
Approach:	They propose a single-sequence prediction method that integrates a graph structure linking key entities in each context passage to relevant subsequent passages for each question.
Outcome:	The proposed method improves answer exact-match/F1 scores and faithfulness of grounding on the hotpotQA dataset and achieves state-of-the-art numbers on the Musique dataset.

Probing LLM World Models: Enhancing Guesstimation with Wisdom of Crowds Decoding (2025.emnlp-main)

Copied to clipboard

Yun-Shiuan Chuang, Sameer Narendran, Nikunj Harlalka, Alexander Cheung, Sizhe Gao, Siddharth Suresh, Junjie Hu, Timothy T. Rogers

Challenge:	a common real-world skill of guesstimation is underexplored in large language model research . a recent study suggests that LLMs encode a world model that supports approximate reasoning .
Approach:	They propose to decode a guesstimation dataset using MARBLES, FUTURE, and ELECPRED . they replicate WOC effects in human participants and find similar benefits .
Outcome:	The proposed model improves accuracy over greedy, self-consistency, and mean decoding in human participants.

Learning to Plan for Retrieval-Augmented Large Language Models from Knowledge Graphs (2024.findings-emnlp)

Copied to clipboard

Junjie Wang, Mingyang Chen, Binbin Hu, Dan Yang, Ziqi Liu, Yue Shen, Peng Wei, Zhiqiang Zhang, Jinjie Gu, Jun Zhou, Jeff Pan, Wen Zhang, Huajun Chen

AfroMT: Pretraining Strategies and Reproducible Benchmarks for Translation of 8 African Languages (2021.emnlp-main)

Challenge:	Recent studies have attempted to enhance the performance of large language models (LLMs) in complex question-answering (QA) tasks by combining step-wise planning with external retrieval.
Approach:	They propose a framework for enhancing LLMs’ planning capabilities by using planning data derived from knowledge graphs (KGs).
Outcome:	The proposed framework improves LLMs’ planning capabilities by using knowledge graphs (KGs) the proposed framework is compared with existing frameworks on multiple datasets and shows that it is effective for large language models.

Copied to clipboard

Machel Reid, Junjie Hu, Graham Neubig, Yutaka Matsuo

Formally Specifying the Intended Behavior of the Program: LLM-Driven Neuro-Symbolic Program Specification Synthesis (2026.acl-demo)

Challenge:	Existing reproducible benchmarks for machine translation are limited to high-resource or well-represented languages.
Approach:	They propose to use AfroMT to develop a reproducible machine translation benchmark for eight widely spoken African languages and a suite of analysis tools to take into account their unique properties.
Outcome:	The proposed benchmarks show significant improvements when pretraining on 11 languages, with gains of up to 2 BLEU points over strong baselines.

Copied to clipboard

Cheng Wen, Hu Junjie, YiKun Hu, Jie Su, Bin Yu, Dugang Liu, Zhiwu Xu, Weidi Sun, Shengchao Qin, Cong Tian

Explicit Alignment Objectives for Multilingual Bidirectional Encoders (2021.naacl-main)

Challenge:	Formal verification typically requires developers to write detailed formal specifications . a formal verification system that generates candidate specifications is costly and error-prone .
Approach:	They propose an LLM-driven neuro-symbolic demonstration system that reframes specification writing as constrained structured synthesis.
Outcome:	The proposed system reduces hallucinations and produces proof-ready annotations.

Copied to clipboard

Junjie Hu, Melvin Johnson, Orhan Firat, Aditya Siddhant, Graham Neubig

Syntactic Text Processing

Challenge:	Pre-trained cross-lingual encoders have proven impressively effective at enabling transfer-learning of NLP systems from high-resource languages to low-resourced languages.
Approach:	They propose a method to align multilingual encoders using two explicit alignment objectives that align the multilingual representations at different granularities.
Outcome:	The proposed method achieves gains of up to 1.1 average F1 score on sequence tagging and 27.3 average accuracy on retrieval over the XLM-R-large model.

Beyond Demographics: Aligning Role-playing LLM-based Agents Using Human Belief Networks (2024.findings-emnlp)

Copied to clipboard

Yun-Shiuan Chuang, Krirk Nirunwiroj, Zach Studdiford, Agam Goyal, Vincent Frigo, Sijia Yang, Dhavan Shah, Junjie Hu, Timothy Rogers

Challenge:	Existing large language models can be prompted to role-play as individuals with particular demographic traits, but results are often human-like.
Approach:	They found that seeding LLM-based agents with a single belief improved alignment . they say that role-playing based on demographic information does not improve alignment a .
Outcome:	The proposed approach improves LLM alignment with human behavior . seeding agents with a single belief improves alignment for topics related to the belief network .

AdapterDistillation: Non-Destructive Task Composition with Knowledge Distillation (2023.emnlp-industry)

Copied to clipboard

Junjie Wang, Yicheng Chen, Wangshu Zhang, Sen Hu, Teng Xu, Jing Zheng

Green & Sustainable NLP

Challenge:	Recent work on learning from multiple tasks has shown that adding an extra fusion layer to implement knowledge composition is non-scalable for some applications.
Approach:	They propose a two-stage knowledge distillation algorithm to extract task specific knowledge by using local data to train a student adapter.
Outcome:	Experiments on frequently asked question retrieval in task-oriented dialog systems validate the efficiency of AdapterDistillation.

MENTOR: Efficient Autoregressive Image Generation with Balanced Multimodal Control (2026.findings-acl)

Copied to clipboard

Haozhe Zhao, Zefan Cai, Shuzheng Si, Liang Chen, Jiuxiang Gu, Wen Xiao, Minjia Zhang, Junjie Hu

Benchmarking Machine Translation with Cultural Awareness (2024.findings-emnlp)

Challenge:	Recent text-to-image models achieve impressive visual quality but still face challenges in precise controllability, balancing multimodal inputs, and high training cost for multimodal image generation.
Approach:	They propose an autoregressive framework with a two-stage training paradigm for controllable multimodal image generation.
Outcome:	Extensive experiments on DreamBench++ and DreamBech show that the proposed framework achieves a strong balance between textual and visual guidance for controllable image generation.

Copied to clipboard

Binwei Yao, Ming Jiang, Tara Bobinac, Diyi Yang, Junjie Hu

Automatic Estimation of Simultaneous Interpreter Performance (P18-2)

Challenge:	Existing studies on terminology translation focus on breaking language barriers rather than cultural barriers.
Approach:	They propose a parallel corpus enriched with CSI annotations in 6 language pairs for investigating Cultural-Aware Machine Translation.
Outcome:	The proposed corpus is enriched with CSI annotations in 6 languages and measures translation quality.

Copied to clipboard

Craig Stewart, Nikolai Vogler, Junjie Hu, Jordan Boyd-Graber, Graham Neubig

Data Augmentation using LLMs: Data Perspectives, Learning Paradigms and Challenges (2024.findings-acl)

Challenge:	Existing methods to predict interpreter confidence and the adequacy of the interpreted message are lacking.
Approach:	They propose to extend a QE pipeline to estimate interpreter performance by using five settings in three language pairs.
Outcome:	The proposed method can predict interpreter confidence and adequacy over five settings in three language pairs and improves interpretation strategy and evaluation measures.

Copied to clipboard

Bosheng Ding, Chengwei Qin, Ruochen Zhao, Tianze Luo, Xinze Li, Guizhen Chen, Wenhan Xia, Junjie Hu, Anh Tuan Luu, Shafiq Joty

Responsible & Trustworthy NLP

Challenge:	Data augmentation (DA) is a key technique for enhancing model performance by diversifying training examples without the need for additional data collection.
Approach:	They examine various strategies that utilize LLMs for data augmentation, including a novel exploration of learning paradigms where LLM-generated data is used for diverse forms of further training.
Outcome:	The proposed approach addresses the primary open challenges faced by LLMs in the field of large language models and aims to serve as a comprehensive guide for researchers and practitioners.

Retrieving Multimodal Prompts for Generative Visual Question Answering (2023.findings-acl)

Copied to clipboard

Timothy Ossowski, Junjie Hu

Domain Differential Adaptation for Neural Machine Translation (D19-56)

Challenge:	Visual question answering (VQA) is a multimodal machine learning problem that challenges a model to answer a question posed about an image.
Approach:	They propose a generative model enhanced by multimodal prompt retrieval that integrates retrieved prompts and multimodal features to generate answers in free text.
Outcome:	The proposed model outperforms its non-retrieval counterpart by 30% on medical VQA tasks.

Copied to clipboard

Zi-Yi Dou, Xinyi Wang, Junjie Hu, Graham Neubig

GRAFF: GRaph-Augmented Fine-grained Fusion for Large Language Models (2026.findings-eacl)

Challenge:	Neural networks are data hungry and domain sensitive, so it is difficult to obtain labeled data for every domain.
Approach:	They propose a framework for domain adaptation where we model the difference between domains instead of smoothing over them.
Outcome:	The proposed framework improves on domain adaptation in multiple experimental settings.

Copied to clipboard

Himanshu Chaudhary, Ruida Wang, Gowtham Ramesh, Junjie Hu

Structured Data in NLP

Challenge:	Existing methods to integrate graphs into LLMs compress the graph's structural information into a single token, restricting their ability to capture deep semantic and structural information.
Approach:	They propose a method that integrates fine-grained node-level structural information with corresponding text entities to LLMs via a lightweight, structure adapter module.
Outcome:	The proposed method outperforms baseline models in graph-based question answering by 10.24%.

Unsupervised Multimodal Neural Machine Translation with Pseudo Visual Pivoting (2020.acl-main)

Copied to clipboard

Po-Yao Huang, Junjie Hu, Xiaojun Chang, Alexander Hauptmann

compare-mt: A Tool for Holistic Comparison of Language Generation Systems (N19-4)

Challenge:	Unsupervised machine translation (MT) has recently achieved impressive results with monolingual corpora.
Approach:	They propose to utilize visual content for disambiguation and promoting latent space alignment in unsupervised machine translation by using multimodal back-translation and pseudo visual pivoting.
Outcome:	The proposed model improves over state-of-the-art methods and generalizes well when images are not available at the testing time.

Copied to clipboard

Graham Neubig, Zi-Yi Dou, Junjie Hu, Paul Michel, Danish Pruthi, Xinyi Wang

Text Generation

Challenge:	Unlike machine translation, natural language outputs are nuanced and there are no clear yes/no distinctions about whether they are correct or not.
Approach:	They describe compare-mt, a tool for holistic analysis and comparison of the results of systems for language generation tasks such as machine translation.
Outcome:	The compare-mt tool is an open-source pure-python package that has already proven useful to generate analyses that have been used in our papers.

EEE-QA: Exploring Effective and Efficient Question-Answer Representations (2024.lrec-main)

Copied to clipboard

Zhanghao Hu, Yijun Yang, Junjie Xu, Yifu Qiu, Pinzhen Chen

GUICourse: From General Vision Language Model to Versatile GUI Agent (2025.acl-long)

Challenge:	Current approaches to question answering rely on pre-trained language models like RoBERTa.
Approach:	They propose a pooling approach that embeds all answer candidates with the question . they also propose enabling cross-reference between answer choices .
Outcome:	The proposed methods improve throughput and memory efficiency with little sacrifice in performance.

Copied to clipboard

Wentong Chen, Junbo Cui, Jinyi Hu, Yujia Qin, Junjie Fang, Yue Zhao, Chongyi Wang, Jun Liu, Guirong Chen, Yupeng Huo, Yuan Yao, Yankai Lin, Zhiyuan Liu, Maosong Sun

Visual Data in NLP

Challenge:	Graphical User Interfaces (GUIs) are a pivotal medium for human-computer interaction.
Approach:	They propose a series of datasets for training visual-based GUI agents using general VLMs.
Outcome:	The proposed GUICourse datasets show that even a small-sized GUI agent performs better on GUI tasks.

How does Multi-Task Training Affect Transformer In-Context Capabilities? Investigations with Function Classes (2024.naacl-short)

Copied to clipboard

Harmon Bhasin, Timothy Ossowski, Yiqiao Zhong, Junjie Hu

Responsible & Trustworthy NLP

Challenge:	Multi-task learning (MTL) for generalist models is a promising direction that offers transfer learning potential.
Approach:	They propose to combine multi-task learning (MTL) with in-context learning (ICL) to build models that can generalize to multiple tasks while being robust to out-of-distribution examples.
Outcome:	The proposed training strategies enable models to learn difficult tasks while mixing in prior tasks, denoted as mixed curriculum.

LegalAgentBench: Evaluating LLM Agents in Legal Domain (2025.acl-long)

Copied to clipboard

Haitao Li, Junjie Chen, Jingli Yang, Qingyao Ai, Wei Jia, Youfeng Liu, Kai Lin, Yueyue Wu, Guozhi Yuan, Yiran Hu, Wuyue Wang, Yiqun Liu, Minlie Huang

Challenge:	Existing general-domain benchmarks do not capture complexity of real-world judicial cognition and decision-making.
Approach:	They propose a benchmark specifically designed to evaluate LLM Agents in the legal domain.
Outcome:	The proposed benchmark includes 17 corpora from real-world legal scenarios and provides 37 tools for interacting with external knowledge.

XTREME-R: Towards More Challenging and Nuanced Multilingual Evaluation (2021.emnlp-main)

Copied to clipboard

Sebastian Ruder, Noah Constant, Jan Botha, Aditya Siddhant, Orhan Firat, Jinlan Fu, Pengfei Liu, Junjie Hu, Dan Garrette, Graham Neubig, Melvin Johnson

Unsupervised Domain Adaptation for Neural Machine Translation with Domain-Aware Feature Embeddings (D19-1)

Challenge:	Recent advances in multilingual natural language processing have improved performance on benchmarks such as XTREME and XGLUE by 13 points . however, improvements have been easier to achieve in some tasks than others .
Approach:	They extend XTREME to XTRAME-R, which includes ten natural language understanding tasks and covers 50 typologically diverse languages.
Outcome:	The proposed framework improves the performance on the XTREME multilingual benchmark by 13 points compared to human-level performance on English transfer learning.

Copied to clipboard

Zi-Yi Dou, Junjie Hu, Antonios Anastasopoulos, Graham Neubig

GroupRank: A Groupwise Paradigm for Effective and Efficient Passage Reranking with LLMs (2026.findings-acl)

Challenge:	Recent studies have focused on domain adaptation for neural machine translation systems where in-domain data is scarce or nonexistent.
Approach:	They propose an approach that adapts models with domain-aware feature embeddings, which are learned via an auxiliary language modeling task.
Outcome:	The proposed model performs better in multiple experimental settings and with back translation.

Copied to clipboard

Meixiu Long, Duolin Sun, Dan Yang, Yihan Jiao, Lei Liu, Jiahai Wang, Binbin Hu, Yue Shen, Jie Feng, Zhehao Tan, Junjie Wang, Lianzhen Zhong, Jian Wang, Peng Wei, Jinjie Gu

Responsible & Trustworthy NLP

Challenge:	Existing rankers excel in lexical-matching scenarios, while they struggle with complex queries requiring deep reasoning.
Approach:	They propose a new paradigm that balances flexibility and context awareness to unlock the full potential of groupwise reranking.
Outcome:	The proposed approach achieves a state-of-the-art 65.2 NDCG@10 on BRIGHT and surpasses baselines by 2.1 points on R2MED while delivering a 6.4 inference speedup.

From Heads to Neurons: Causal Attribution and Steering in Multi-Task Vision–Language Models (2026.findings-acl)

Copied to clipboard

Qidong Wang, Junjie Hu, Ming Jiang

Language Models

Challenge:	Existing models focus on single tasks, limiting comparability of neuron importance . ranking strategies overlook how task-dependent information pathways shape write-in effects of feed-forward network (FFN) neurons.
Approach:	They propose a gradient-free framework for task-aware neuron attribution and steering in multi-task vision-language models.
Outcome:	The proposed framework outperforms existing methods in identifying task-critical neurons and improves model performance after steering.

Learning Label Hierarchy with Supervised Contrastive Learning (2024.findings-eacl)

Copied to clipboard

Ruixue Lian, William Sethares, Junjie Hu

Domain Adaptation of Neural Machine Translation by Lexicon Induction (P19-1)

Challenge:	Existing approaches to supervised contrastive learning treat each class as independent and therefore consider all classes to be equally important.
Approach:	They propose a family of Label-Aware SCL methods that incorporate hierarchical information to SCL by leveraging similarities between classes.
Outcome:	The proposed method outperforms baseline supervised approaches on three datasets.

Copied to clipboard

Junjie Hu, Mengzhou Xia, Graham Neubig, Jaime Carbonell

Text Generation

Challenge:	Neural machine translation (NMT) is sensitive to domain shift, resulting in failure for sentences with large numbers of unknown words and lack of supervision for domain-specific words.
Approach:	They propose an unsupervised method which fine-tunes a pre-trained out-of-domain NMT model using a pseudo-in-domain corpus.
Outcome:	The proposed method improves in five domains without using in-domain parallel sentences and up to 2 BLEU over strong back-translation baselines.

V-SEAM: Visual Semantic Editing and Attention Modulating for Causal Interpretability of Vision-Language Models (2025.emnlp-main)

Copied to clipboard

Qidong Wang, Junjie Hu, Ming Jiang

Explainability & Interpretability in NLP

Challenge:	Existing work on causal interpretability focuses on large language models (LLMs) but internal mechanisms of vision-language models remain underexplored, authors say .
Approach:	They introduce a framework that combines visual and semantic manipulations for causal interpretation of vision-language models.
Outcome:	The proposed framework shows improved performance for LLAVA and InstructBLIP on three diverse benchmarks.

Rapid Adaptation of Neural Machine Translation to New Languages (D18-1)

Copied to clipboard

Graham Neubig, Junjie Hu

DEEP: DEnoising Entity Pre-training for Neural Machine Translation (2022.acl-long)

Challenge:	Existing approaches to adapt neural machine translation systems to low-resource languages are difficult to implement and require large amounts of training data.
Approach:	They propose a method to train neural machine translation systems to new low-resource languages . they propose to start with massively multilingual "seed models" and continue training on data related to the LRL .
Outcome:	The proposed method achieves BLEU scores of up to 15.5 with no data from the LRL and improves over other adaptation methods by 1.7 BLUE points average over 4 LRL settings.

Copied to clipboard

Junjie Hu, Hiroaki Hayashi, Kyunghyun Cho, Graham Neubig

Local Byte Fusion for Neural Machine Translation (2023.acl-long)

Challenge:	Earlier named entity translation methods focus on phonetic transliteration, which ignores the sentence context for translation.
Approach:	They propose a DEnoising Entity Pre-training method that leverages monolingual data and a knowledge base to improve named entity translation accuracy within sentences.
Outcome:	The proposed method improves on three language pairs and denoising auto-encoding baselines.

Copied to clipboard

Makesh Narsimhan Sreedhar, Xiangpeng Wan, Yu Cheng, Junjie Hu

Is Fine-tuning Needed? Pre-trained Language Models Are Near Perfect for Out-of-Domain Detection (2023.acl-long)

Challenge:	Existing NLP models rely on a pre-built subword tokenizer to tokenize a sentence . this can be rigid and subwords from low-resource languages are under-represented .
Approach:	They propose a method for byte-based machine translation that aggregates local semantic information.
Outcome:	The proposed method improves on multilingual translation and cross-lingual transfer . it is parameter-efficient and performs competitively to subword models, it is shown .

Copied to clipboard

Rheeya Uppaal, Junjie Hu, Yixuan Li

Beyond Counting Datasets: A Survey of Multilingual Dataset Construction and Necessary Resources (2022.findings-emnlp)

Challenge:	Out-of-distribution (OOD) detection is critical for reliable predictions over text . fine-tuning with pre-trained language models has been a de facto procedure .
Approach:	They propose to leverage pre-trained language models for OOD detection without fine-tuning on ID data.
Outcome:	The proposed approach outperforms the fine-tuned model under distributional shifts.

Copied to clipboard

Xinyan Yu, Trina Chatterjee, Akari Asai, Junjie Hu, Eunsol Choi

REO-Relevance, Extraness, Omission: A Fine-grained Evaluation for Image Captioning (D19-1)

Challenge:	Existing studies have examined the quality of labeled data in non-English languages.
Approach:	They annotate how datasets are created, input text and label sources, tools used to build them and what they study.
Outcome:	The results show that language-proficient NLP researchers' estimated availability correlates with dataset availability.

Copied to clipboard

Ming Jiang, Junjie Hu, Qiuyuan Huang, Lei Zhang, Jana Diesner, Jianfeng Gao

Simulating Opinion Dynamics with Networks of LLM-based Agents (2024.findings-naacl)

Challenge:	Existing metrics for image captioning evaluation provide an overall quality score, which is difficult to infer specific description errors.
Approach:	They propose a fine-grained evaluation method REO for automatically measuring the performance of image captioning systems.
Outcome:	The proposed method achieves higher consistency with human judgments and provides more intuitive evaluation results than other metrics.

Copied to clipboard

Yun-Shiuan Chuang, Agam Goyal, Nikunj Harlalka, Siddharth Suresh, Robert Hawkins, Sijia Yang, Dhavan Shah, Junjie Hu, Timothy Rogers

Challenge:	Existing approaches to simulating opinion dynamics often over-simplify human behavior . authors propose refining LLMs with real-world discourse to better simulate evolution of beliefs .
Approach:	They propose to use large language models to simulate opinion dynamics in groups of simulated agents . they found that LLM agents produce more accurate information than ABMs .
Outcome:	The proposed model can be used to better simulate opinion dynamics in real-world discourses.

OLIVE: Object Level In-Context Visual Embeddings (2024.acl-long)

Copied to clipboard

Timothy Ossowski, Junjie Hu

Utilizing Language-Image Pretraining for Efficient and Robust Bilingual Word Alignment (2022.findings-emnlp)

Challenge:	Existing vision-language models lack fine-grained object-level understanding and grounding . existing models implicitly align text tokens with image patch tokens, which is ineffective for embedding alignment at the same granularity and introduces noisy spurious background features.
Approach:	They propose a method to prompt large language models with in-context visual object vectors . this method allows for controllable object-level reasoning .
Outcome:	The proposed method achieves competitive referring object classification and captioning performance while offering zero-shot generalization and robustness to visually challenging contexts.

Copied to clipboard

Tuan Dinh, Jy-yong Sohn, Shashank Rajput, Timothy Ossowski, Yifei Ming, Junjie Hu, Dimitris Papailiopoulos, Kangwook Lee