Papers by Julian McAuley
Interview: Large-scale Modeling of Media Dialog with Discourse Patterns and Knowledge Grounding (2020.emnlp-main)
Copied to clipboard
| Challenge: | Discourse analysis has been limited to small news corpora, but this study is expanding to tens of thousands of interviews. |
| Approach: | They propose a large-scale analysis of discourse in media dialog and its impact on dialog modeling with a focus on interrogative patterns and use of external knowledge. |
| Outcome: | The proposed model outperforms strong discourse-agnostic baselines for dialog modeling, generating more specific and topical responses in interview-style conversations. |
Unsupervised Improvement of Factual Knowledge in Language Models (2023.eacl-main)
Copied to clipboard
| Challenge: | Masked language modeling (MLM) is often dominated by high-frequency words that are sub-optimal for learning factual knowledge. |
| Approach: | They propose an approach that forces the model to prioritize informative words in a fully unsupervised way. |
| Outcome: | The proposed approach significantly improves the performance of pretrained language models on factual recall, question answering, sentiment analysis, and natural language inference in a closed-book setting. |
Few-shot Dialogue Strategy Learning for Motivational Interviewing via Inductive Reasoning (2024.findings-acl)
Copied to clipboard
Zhouhang Xie, Bodhisattwa Prasad Majumder, Mengjie Zhao, Yoshinori Maeda, Keiichi Yamada, Hiromi Wakaki, Julian McAuley
| Challenge: | Motivational Interviewing (MI) requires a system that can infer how to motivate users to adopt positive lifestyle changes. |
| Approach: | They propose a framework that can learn and apply conversation strategies from expert demonstrations by using natural language inductive rules. |
| Outcome: | The proposed framework outperforms in-context demonstrations that are over 50 times longer and can learn natural language strategies from demonstrations. |
SkipBERT: Efficient Inference with Shallow Layer Skipping (2022.acl-long)
Copied to clipboard
| Challenge: | Pre-trained language models have significant demands in computation and inference time, limiting their use in resource-constrained or latencysensitive applications. |
| Approach: | They propose to encode text chunks into independent representations and skip computation of shallow layers to accelerate inference. |
| Outcome: | The proposed approach can reduce latency by 65% without sacrificing performance. |
LaPraDoR: Unsupervised Pretrained Dense Retriever for Zero-Shot Text Retrieval (2022.findings-acl)
Copied to clipboard
| Challenge: | Experimental results show that LaPraDoR is state-of-the-art compared with supervised dense retrieval models. |
| Approach: | They propose a pretrained dual-tower dense retriever that does not require supervised data for training. |
| Outcome: | The proposed method achieves state-of-the-art performance on 18 datasets of 9 zero-shot text retrieval tasks. |
Unlocking Decoding-time Controllability: Gradient-Free Multi-Objective Alignment with Contrastive Prompts (2025.naacl-long)
Copied to clipboard
| Challenge: | Existing methods for aligning large language models with human preferences are poor in extensibility and require significant retraining. |
| Approach: | They propose a multi-objective alignment approach that constructs an expert prompt and an adversarial prompt for each alignment objective to contrast at the decoding time. |
| Outcome: | The proposed approach is superior to existing methods in obtaining a well-distributed Pareto front among different alignment objectives. |
Doc-React: Multi-page Heterogeneous Document Question-answering (2025.acl-short)
Copied to clipboard
Junda Wu, Yu Xia, Tong Yu, Xiang Chen, Sai Sree Harsha, Akash V Maharaj, Ruiyi Zhang, Victor Bursztyn, Sungchul Kim, Ryan A. Rossi, Julian McAuley, Yunyao Li, Ritwik Sinha
| Challenge: | Existing methods for integrating information across multiple modalities are suboptimal for multi-page, multimodal documents. |
| Approach: | They propose an adaptive iterative framework that balances information gain and uncertainty reduction at each step. |
| Outcome: | The proposed framework captures relevant multimodal content and achieves strong performance on complex QA tasks. |
CachePrune: Teaching LLMs What Not to Follow via KV-Cache Editing (2026.acl-long)
Copied to clipboard
Rui Wang, Junda Wu, Yu Xia, Tong Yu, Ruiyi Zhang, Ryan A. Rossi, Subrata Mitra, Lina Yao, Julian McAuley
| Challenge: | Existing Large Language Models exhibit critical vulnerability to indirect prompt injection attacks, where instructions injected within in the prompt context can override the user's intent. |
| Approach: | They propose a neural pruning algorithm that prunes neurons associated with instruction-following during KV cache encoding of the prompt context. |
| Outcome: | The proposed approach significantly reduces the attack success rate while preserving the model's ability to follow user instructions. |
Explainable Chain-of-Thought Reasoning: An Empirical Analysis on State-Aware Reasoning Dynamics (2025.findings-emnlp)
Copied to clipboard
Sheldon Yu, Yuxin Xiong, Junda Wu, Xintong Li, Tong Yu, Xiang Chen, Ritwik Sinha, Jingbo Shang, Julian McAuley
| Challenge: | Recent advances in chain-of-thought prompting have demonstrated the ability of large language models to perform multi-step reasoning. |
| Approach: | They propose a framework to analyze latent dynamics of CoT trajectories for interpretability . they segment generated CoT into discrete reasoning steps and abstract each step into a spectral embedding based on token-level Gram matrices . |
| Outcome: | The proposed framework segments generated CoT steps into discrete reasoning steps, abstracts each step into a spectral embedding based on token-level Gram matrices, and clusters these embeddements into semantically meaningful latent states. |
Mitigating Visual Knowledge Forgetting in MLLM Instruction-tuning via Modality-decoupled Gradient Descent (2025.findings-emnlp)
Copied to clipboard
Junda Wu, Yuxin Xiong, Xintong Li, Yu Xia, Ruoyu Wang, Yu Wang, Tong Yu, Sungchul Kim, Ryan A. Rossi, Lina Yao, Jingbo Shang, Julian McAuley
| Challenge: | Existing fine-tuning and continual learning methods compress visual representations and emphasize task alignment over visual retention. |
| Approach: | They propose a modality-decoupled gradient descent (MDGD) that regulates gradient updates to preserve effective rank of visual features and explicitly disentangles visual learning from task-specific alignment. |
| Outcome: | The proposed model reduces visual forgetting and improves visual retention . it disentangles visual learning from task-specific alignment and preserves effective rank . |
Zero-shot Generalization in Dialog State Tracking through Generative Question Answering (2021.eacl-main)
Copied to clipboard
| Challenge: | Existing methods for Dialog State Tracking do not generalize well to new domains and unseen slots. |
| Approach: | They propose an ontology-free framework that queries for unseen constraints and slots in multi-domain task-oriented dialogs using a conditional language model pre-trained on substantive English sentences. |
| Outcome: | The proposed framework improves goal accuracy in zero-shot domain adaptation settings by up to 9% over the previous state-of-the-art on the MultiWOZ 2.1 dataset. |
Detect and Perturb: Neutral Rewriting of Biased and Sensitive Text via Gradient-based Decoding (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Written language carries explicit and implicit biases that can distract from meaningful signals; at worst they can lead to unfair outcomes. |
| Approach: | They propose a gradient-based rewriting framework that detects and perturbs sensitive components and regenerates fluent alternatives that are neutral in the sensitive attribute while maintaining the semantics of other attributes. |
| Outcome: | The proposed framework regenerates fluent alternatives that are neutral in the sensitive attribute while maintaining the semantics of other attributes. |
Beyond Preserved Accuracy: Evaluating Loyalty and Robustness of BERT Compression (2021.emnlp-main)
Copied to clipboard
| Challenge: | Recent studies on compression of pretrained language models usually use preserved accuracy as the metric for evaluation. |
| Approach: | They propose two new metrics that measure how closely a compressed model mimics the original model. |
| Outcome: | The proposed metrics measure how closely a compressed model (i.e., student) mimics the original model (e.g., teacher). |
Baize: An Open-Source Chat Model with Parameter-Efficient Tuning on Self-Chat Data (2023.emnlp-main)
Copied to clipboard
| Challenge: | Despite the promising potential of chat models, they are only accessible through restricted APIs, creating barriers for new research and progress in the field. |
| Approach: | They propose a pipeline that can automatically generate a high-quality multi-turn chat corpus by leveraging ChatGPT to engage in a conversation with itself. |
| Outcome: | The proposed pipeline generates a high-quality multi-turn chat corpus by leveraging ChatGPT to engage in a conversation with itself, simulating both user and AI responses. |
Image Difference Captioning via Adversarial Preference Optimization (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing supervised approaches to image difference captioning overfit to dataset-specific language patterns and fail to capture accurate preferences. |
| Approach: | They propose an adversarial direct preference optimization framework that aligns captioning policy with pairwise difference preferences via Direct Preference Optimization. |
| Outcome: | The proposed approach outperforms baselines on benchmark IDC datasets in generating fine-grained and accurate difference descriptions. |
Scalable and Accurate Dialogue State Tracking via Hierarchical Sequence Generation (D19-1)
Copied to clipboard
| Challenge: | Existing approaches to dialogue state tracking rely on pre-defined ontologies . however, these methods suffer from computational complexity that increases proportionally to the number of pre-determined slots. |
| Approach: | They propose a model that generates a sequence of belief states without the pre-defined ontology list. |
| Outcome: | The proposed model scales easily with the increasing number of pre-defined slots and domains and reaches the state-of-the-art performance on the multi-domain and single domain dialogue state tracking datasets. |
Blow the Dog Whistle: A Chinese Dataset for Cant Understanding with Common Sense and World Knowledge (2021.naacl-main)
Copied to clipboard
| Challenge: | Cant is important for understanding advertising, comedies and dogwhistle politics . currently, there are very few resources available for the research of cant . |
| Approach: | They propose a large and diverse dataset for creating and understanding cant from a computational linguistics perspective. |
| Outcome: | The proposed dataset can be used to test word embedding similarity and pretrained language models. |
BERT Learns to Teach: Knowledge Distillation with Meta Learning (2022.acl-long)
Copied to clipboard
| Challenge: | Existing knowledge distillation methods are based on teacher model, but have drawbacks . a teacher model is fixed during training, but meta learning can improve student performance . |
| Approach: | They propose a meta learning framework that allows the teacher network to learn to better transfer knowledge to the student network. |
| Outcome: | Experiments show that MetaDistil can improve on existing methods and is less sensitive to student capacity and hyperparameters. |
Extending Input Contexts of Language Models through Training on Segmented Sequences (2024.findings-naacl)
Copied to clipboard
| Challenge: | Effectively training languages models on long sequences poses many technical challenges. |
| Approach: | They propose a method for extending positional embeddings by sub-sampling segments from long inputs while maintaining their original position. |
| Outcome: | The proposed method extends the input con-text size of pretrained models without any changes in the model's memory and memory costs. |
Personalized Review Generation By Expanding Phrases and Attending on Aspect-Aware Representations (P18-2)
Copied to clipboard
| Challenge: | Existing systems that use user and item identity as inputs for review generation are lacking in the field of natural language processing. |
| Approach: | They propose an encoder-decoder framework that generates personalized reviews by expanding short phrases provided as input to the system. |
| Outcome: | The proposed model learns representations capable of generating coherent and diverse reviews. |
Synthetic Pre-Training Tasks for Neural Machine Translation (2023.findings-acl)
Copied to clipboard
| Challenge: | toxicity and bias can be addressed by pre-training with synthetic resources . BLEU scores are used to compare methods with real-world data . |
| Approach: | They propose several ways to generate obfuscated data from large parallel corpus and concatenating phrase pairs from small word-aligned corpus with synthetic parallel data without real human language corpora. |
| Outcome: | The proposed methods can be used to generate obfuscated data or synthetic parallel data without real human language corpora even with high levels of oblication. |
Spoiler Detection as Semantic Text Matching (2023.emnlp-main)
Copied to clipboard
| Challenge: | Existing research on spoiler detection shows promising results in safeguarding viewers from general spoilers, but it fails to address the issue of users abstaining from show-related content during their watch. |
| Approach: | They propose to use semantic text matching to assign an episode number to a spoiler given a specific TV show and a dataset to evaluate its performance. |
| Outcome: | The proposed dataset can be used to evaluate the performance of the proposed model and to compare it with other datasets. |
Bridging Language and Items for Retrieval and Recommendation: Benchmarking LLMs as Semantic Encoders (2026.acl-long)
Copied to clipboard
| Challenge: | Recent advances in large language models have enabled their use as semantic encoders for recommendation, but their roles and behaviors in this setting are still not well understood. |
| Approach: | They propose a benchmark to evaluate large language models as semantic encoders in recommendation scenarios. |
| Outcome: | The proposed benchmark shows that ranking of 11 leading LLMs is low compared to MTEB, highlighting the unique challenges of semantic encoding in recommendation. |
From Selection to Generation: A Survey of LLM-based Active Learning (2025.acl-long)
Copied to clipboard
Yu Xia, Subhojyoti Mukherjee, Zhouhang Xie, Junda Wu, Xintong Li, Ryan Aponte, Hanjia Lyu, Joe Barrow, Hongjie Chen, Franck Dernoncourt, Branislav Kveton, Tong Yu, Ruiyi Zhang, Jiuxiang Gu, Nesreen K. Ahmed, Yu Wang, Xiang Chen, Hanieh Deilamsalehy, Sungchul Kim, Zhengmian Hu, Yue Zhao, Nedim Lipka, Seunghyun Yoon, Ting-Hao Kenneth Huang, Zichao Wang, Puneet Mathur, Soumyabrata Pal, Koyel Mukherjee, Zhehao Zhang, Namyong Park, Thien Huu Nguyen, Jiebo Luo, Ryan A. Rossi, Julian McAuley
| Challenge: | Large Language Models (LLMs) have been used for selection and training of data for active learning. |
| Approach: | They propose an intuitive taxonomy that categorizes LLM-based active learning techniques and discuss the transformative roles they can play in the active learning loop. |
| Outcome: | The proposed model can generate entirely new data instances and provide more cost-effective annotations with fewer labeled data instances. |
Linear Layer Extrapolation for Fine-Grained Emotion Classification (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing studies show that Transformer-based language models are more factual accurate in later layers . |
| Approach: | They propose a method that optimizes contrast based on the selected intermediate layer . they observe a similar pattern for fine-grained emotion classification in text . |
| Outcome: | Experiments show that the proposed method outperforms standard methods in fine-grained emotion classification tasks. |
KNOW How to Make Up Your Mind! Adversarially Detecting and Alleviating Inconsistencies in Natural Language Explanations (2023.acl-short)
Copied to clipboard
| Challenge: | eIA is an adversarial attack that generates inconsistent natural language explanations (NLEs) a model that generate In-NLE is undesirable, as it has a faulty decision-making process or is prone to inconsistencies. |
| Approach: | They propose an off-the-shelf mitigation method to alleviate inconsistencies by grounding the model into external background knowledge. |
| Outcome: | The proposed method reduces inconsistencies detected by previous models . it is based on external knowledge bases and a novel approach to mitigate inconsistent models based upon the proposed method . |
Controlling Bias Exposure for Fair Interpretable Predictions (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Existing approaches to reduce bias in NLP tasks focus on protecting or isolating information related to a sensitive attribute, but they lack control over how much bias is required to be removed. |
| Approach: | They propose a favorable debiasing method that uses sensitive information ‘fairly’, rather than blindly eliminating it. |
| Outcome: | The proposed method achieves a trade-off between debiasing and task performance along with producing debiased rationales as evidence. |
A Survey on Dynamic Neural Networks for Natural Language Processing (2023.findings-eacl)
Copied to clipboard
| Challenge: | Dynamic neural networks can scale up pretrainable models with sub-linear increases in computation and time. |
| Approach: | They summarize the progress of three types of dynamic neural networks in NLP . skimming, mixtures of experts, and early exit are among the most popular . |
| Outcome: | The proposed models can scale up with sub-linear increases in computation and time . skimming, mixture of experts, and early exit are the most popular approaches . |
MedEval: A Multi-Level, Multi-Task, and Multi-Domain Medical Benchmark for Language Model Evaluation (2023.emnlp-main)
Copied to clipboard
| Challenge: | Existing medical datasets require high quality domain-specific datasets. |
| Approach: | They propose a multi-level, multi-task, and multi-domain medical benchmark to facilitate the development of language models for healthcare. |
| Outcome: | The proposed model provides granular potential usage and supports a wide range of tasks. |
Efficiently Tuned Parameters Are Task Embeddings (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for intermediate-task transfer are computationally infeasible to experiment with all intermediate combinations. |
| Approach: | They propose to use task-specific parameters updated in parameter-efficient tuning methods to predict inter-task transferability. |
| Outcome: | The proposed approach outperforms existing methods while being conceptually simple and computationally efficient. |
InforMask: Unsupervised Informative Masking for Language Model Pretraining (2022.emnlp-main)
Copied to clipboard
| Challenge: | Masked language modeling is used for pretraining large language models for knowledge-intensive tasks. |
| Approach: | They propose an unsupervised masking strategy that exploits Pointwise Mutual Information to select the most informative tokens to mask. |
| Outcome: | The proposed strategy outperforms random masking and previously proposed masking strategies on the factual recall benchmark LAMA and the question answering benchmark SQuAD v1 and v2. |
Large Language Models and Causal Inference in Collaboration: A Comprehensive Survey (2025.findings-naacl)
Copied to clipboard
Xiaoyu Liu, Paiheng Xu, Junda Wu, Jiaxin Yuan, Yifan Yang, Yuhang Zhou, Fuxiao Liu, Tianrui Guan, Haoliang Wang, Tong Yu, Julian McAuley, Wei Ai, Furong Huang
| Challenge: | Large Language Models (LLMs) have shown great potential to enhance Natural Language Processing (NLP) models in areas such as predictive accuracy, fairness, robustness, and explainability. |
| Approach: | They evaluate or improve generative Large Language Models from a causal perspective in areas such as reasoning capacity, fairness and safety issues, explainability, and handling multimodality. |
| Outcome: | The proposed models can be used to perform causal relationship discovery and causal effect estimation tasks. |
Beyond Chain-of-Thought: A Survey of Chain-of-X Paradigms for LLMs (2025.coling-main)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have shown impressive reasoning abilities when prompted with Chain-of-Thought (CoT). |
| Approach: | They propose to categorize Chain-of-X methods by taxonomies of nodes, i.e., the X in CoX, and application tasks, and then categorise them by taxanomies and discuss potential future directions. |
| Outcome: | The proposed methods are categorised by taxonomies of nodes, i.e., the X in CoX, and application tasks. |
SHARE: a System for Hierarchical Assistive Recipe Editing (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing recipe websites do not provide options for users with dietary restrictions . a growing population follows some form of dietary restriction, with many people following it for a variety of reasons . |
| Approach: | They propose a system for hierarchical assistive recipe editing that performs simultaneous ingredient substitution before generating natural-language steps using the edited ingredients. |
| Outcome: | The proposed system can adapt a recipe to satisfy a user-specified dietary constraint. |
When Benchmarks Age: Temporal Misalignment through Large Language Model Factuality Evaluation (2026.eacl-short)
Copied to clipboard
| Challenge: | Existing studies on LLM factuality evaluation have not investigated the reliability of static evaluation benchmarks. |
| Approach: | They examine five popular factuality benchmarks and eight LLMs released over different years to assess their reliability. |
| Outcome: | The proposed method compared five popular factuality benchmarks and eight LLMs released over different years. |
Assistive Recipe Editing through Critiquing (2023.eacl-main)
Copied to clipboard
| Challenge: | Existing methods for generating recipes that satisfy dietary restrictions are inconsistent or incoherent and paired datasets are not available at scale. |
| Approach: | They propose to build a hierarchical denoising auto-encoder that edits recipes given ingredient-level critiques by interacting with the predicted ingredients. |
| Outcome: | The proposed model can more effectively edit recipes compared to strong language models and iteratively rewrites recipes to satisfy user feedback. |
Fine-Grained Spoiler Detection from Large-Scale Review Corpora (P19-1)
Copied to clipboard
| Challenge: | 'Spoilers' on review websites can be a concern for consumers who want to fully experience the excitement of media consumption. |
| Approach: | They propose to use a large-scale book review dataset to generate fine-grained spoiler annotations . they then use supervised neural networks to detect spoiler sentences in review corpora . |
| Outcome: | The proposed method outperforms baselines in a large-scale book review dataset . it can detect spoiler sentences in review corpora, but only a few users use it . |
WildScore: Benchmarking MLLMs in-the-Wild Symbolic Music Reasoning (2025.emnlp-main)
Copied to clipboard
| Challenge: | Recent advances in Multimodal Large Language Models (MLLMs) have demonstrated impressive capabilities across various vision-language tasks. |
| Approach: | They propose a systematic taxonomy to evaluate MLLMs' ability to interpret real-world music scores and answer complex musicological queries. |
| Outcome: | The proposed model is based on real-world music scores and user-generated questions and discussions, and is scalable and controlled. |
Unsupervised Enrichment of Persona-grounded Dialog with Background Stories (2021.acl-short)
Copied to clipboard
| Challenge: | Existing dialog models do not contain such narratives, so we propose a gradient-based rewriting technique to enrich dialog personas with relevant background events. |
| Approach: | They propose to use existing dialog datasets to enrich dialog responses with 'background stories' based on a gradient-based rewriting technique which encourages the generated response to be fluent with the dialog history, minimally different from the retrieved story, and consistent with the original persona. |
| Outcome: | The proposed method generates responses that are more diverse and human-like compared to outputs from existing dialog models. |
Cognitive Bias in Decision-Making with LLMs (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Large language models inherit societal biases against protected groups and can be subject to functionally resembling cognitive bias. |
| Approach: | They propose a framework to uncover, evaluate, and mitigate cognitive bias in large language models by using a dataset containing 13,465 prompts to evaluate LLM decisions on different cognitive biases. |
| Outcome: | The proposed framework uncovers, evaluates, and mitigates cognitive bias in large language models, particularly in high-stakes decision-making tasks. |
Aligning as Debiasing: Causality-Aware Alignment via Reinforcement Learning with Interventional Feedback (2024.naacl-long)
Copied to clipboard
| Challenge: | Existing methods to reduce LLMs' biased outputs rely on reward signals from current model outputs without considering the source of biases. |
| Approach: | They propose to leverage the reward model in RL alignment as an instrumental variable to perform causal intervention on LLMs. |
| Outcome: | The proposed method reduces biases by using human feedback to fine tune LLMs to human values. |
Automatic Multi-Label Prompting: Simple and Interpretable Few-Shot Classification (2022.naacl-main)
Copied to clipboard
| Challenge: | Prompt-based learning is an emerging paradigm for exploiting knowledge learned by a pretrained language model. |
| Approach: | They propose a method to automatically select label mappings for few-shot text classification with prompting. |
| Outcome: | The proposed method achieves competitive performance on the GLUE benchmark without human effort or external resources. |
Weakly Supervised Contrastive Learning for Chest X-Ray Report Generation (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Radiology report generation aims at generating descriptive text from radiology images automatically. |
| Approach: | They propose a weakly supervised contrastive loss method that generates descriptive text from radiology images automatically. |
| Outcome: | The proposed method outperforms previous work on correctness and text generation metrics for two public benchmarks. |
Embedding-Informed Adaptive Retrieval-Augmented Generation of Large Language Models (2025.coling-main)
Copied to clipboard
| Challenge: | Retrieval-augmented large language models excel in various NLP tasks but are not always helpful when the knowledge required is absent in the model. |
| Approach: | They propose to determine whether the model is knowledgeable on a query via inspecting the (contextualized) pre-trained token embeddings of LLMs. |
| Outcome: | Experiments show that the proposed approach performs better than previous approaches on various benchmarks. |
Automatic Pair Construction for Contrastive Post-training (2024.findings-naacl)
Copied to clipboard
Canwen Xu, Corby Rosset, Ethan Chau, Luciano Corro, Shweti Mahajan, Julian McAuley, Jennifer Neville, Ahmed Awadallah, Nikhil Rao
| Challenge: | Large language models (LLMs) have unprecedented proficiency in a wide array of tasks. |
| Approach: | They propose a way to construct contrastive data using preference pairs from multiple models of varying strengths using SLiC and DPO. |
| Outcome: | The proposed method outperforms existing models like Orca in the comparison of SLiC and DPO with SFT baselines. |
Train Once, Deploy Anywhere: Matryoshka Representation Learning for Multimodal Recommendation (2024.findings-emnlp)
Copied to clipboard
| Challenge: | integrating rich multimodal knowledge into recommender systems remains a challenge . despite performance improvements, different recommendation scenarios often require varying granularities. |
| Approach: | They propose a framework that captures item features at different granularities and learns informative representations for efficient recommendation across multiple dimensions. |
| Outcome: | The proposed framework achieves superior performance over state-of-the-art models on multiple benchmark datasets. |
InterFair: Debiasing with Natural Language Feedback for Fair Interpretable Predictions (2023.emnlp-main)
Copied to clipboard
| Challenge: | Debiasing methods in NLP models focus on isolating information related to a sensitive attribute (e.g., gender or race) but instead argue that a favorable debiaser should use sensitive information ‘fairly,’ with explanations, rather than blindly eliminating it. |
| Approach: | They propose that a favorable debiasing method should use sensitive information ‘fairly,’ with explanations, rather than blindly eliminating it. |
| Outcome: | The proposed approach reduces bias in explanations while maintaining the same prediction accuracy. |
Code to Think, Think to Code: A Survey on Code-Enhanced Reasoning and Reasoning-Driven Code Intelligence in LLMs (2025.emnlp-main)
Copied to clipboard
Dayu Yang, Tianyang Liu, Daoan Zhang, Antoine Simoulin, Xiaoyi Liu, Yuwei Cao, Zhaopu Teng, Xin Qian, Grey Yang, Jiebo Luo, Julian McAuley
| Challenge: | Recent breakthrough models like OpenAI-o1 and DeepSeek-R1 show powerful task-solving capabilities, particularly advances in reasoning. |
| Approach: | They propose future research directions that may deepen the synergy, ultimately advancing LLM performance in both complex reasoning and code intelligence. |
| Outcome: | The proposed research may deepen the synergy, ultimately advancing LLM performance in both complex reasoning and code intelligence. |
CoMMIT: Coordinated Multimodal Instruction Tuning (2025.emnlp-main)
Copied to clipboard
Xintong Li, Junda Wu, Tong Yu, Rui Wang, Yu Wang, Xiang Chen, Jiuxiang Gu, Lina Yao, Julian McAuley, Jingbo Shang
| Challenge: | et al., 2024) show that multimodal instruction tuning is more effective than baselines. |
| Approach: | They propose a multimodal balance coefficient that enables quantitative measurement of the balance of learning . they propose auxiliary regularization on the gradient to promote updating with larger step sizes . |
| Outcome: | The proposed method is more effective than baselines in MLLM instruction tuning. |
UCTopic: Unsupervised Contrastive Learning for Phrase Representations and Topic Mining (2022.acl-long)
Copied to clipboard
| Challenge: | Existing phrase representation learning methods combine unigram representations in a context-free manner or rely on extensive annotations to learn context-aware knowledge. |
| Approach: | They propose a novel unsupervised contrastive learning framework for context-aware phrase representations and topic mining. |
| Outcome: | The proposed framework outperforms the state-of-the-art phrase representation model by 38.2% NMI on four entity clustering tasks. |
Small Models are Valuable Plug-ins for Large Language Models (2024.findings-acl)
Copied to clipboard
| Challenge: | Large-scale pre-trained language models are difficult to fine-tune due to their huge weights and limited context length. |
| Approach: | They propose an approach which allows black-box LLMs to work with locally fine-tuned smaller models, resulting in superior performance on supervised tasks. |
| Outcome: | The proposed approach overcomes the challenges of poor performance and instability of In-Context Learning (ICL) while reducing the complexity of in-context learning. |
Improving Neural Story Generation by Targeted Common Sense Grounding (D19-1)
Copied to clipboard
| Challenge: | Recent advances in language modeling have yielded thematic and stylistic coherence in story generation through large scale pretraining of Transformer models. |
| Approach: | They propose a multi-task learning scheme to achieve better common sense reasoning in language models by leveraging auxiliary training signals from datasets designed to provide common sense grounding. |
| Outcome: | The proposed model achieves improved common sense reasoning and state-of-the-art perplexity on the WritingPrompts dataset. |
Learning to Attend On Essential Terms: An Enhanced Retriever-Reader Model for Open-domain Question Answering (N19-1)
Copied to clipboard
| Challenge: | Existing approaches to open-domain question answering struggle to retrieve indirectly related evidence when no direct evidence is provided. |
| Approach: | They propose a retriever-reader model that learns to attend on essential terms during the question answering process. |
| Outcome: | The proposed model achieves the state-of-the-art on multiple open-domain QA datasets and achieves a 'reader-reader' level. |
Instilling Type Knowledge in Language Models via Multi-Task QA (2022.findings-naacl)
Copied to clipboard
| Challenge: | Current methods to learn entity types rely on coarse, noisy labels . current methods rely only on text-to-text pre-training on type-centric questions . |
| Approach: | They propose to instill fine-grained type knowledge in language models by pre-training on type-centric questions. |
| Outcome: | The proposed model achieves state-of-the-art in zero-shot dialog state tracking benchmarks and can accurately infer entity types in Wikipedia articles. |
InstructGraph: Boosting Large Language Models via Graph-centric Instruction Tuning and Preference Alignment (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing large language models (LLMs) can solve graph reasoning and generation tasks with parameter updates without sacrificing performance. |
| Approach: | They propose a structured format verbalizer to unify all graph data into a universal code-like format, which can simply represent the graph without any external graph-specific encoders. |
| Outcome: | The proposed framework outperforms GPT-4 and LLaMA2 in graph reasoning and generation tasks by more than 13% and 38%, respectively. |
Like hiking? You probably enjoy nature: Persona-grounded Dialog with Commonsense Expansions (2020.emnlp-main)
Copied to clipboard
| Challenge: | Existing persona-grounded dialog models fail to capture simple implications of given persona descriptions. |
| Approach: | They propose to expand available persona sentences using existing commonsense knowledge bases and paraphrasing resources to imbue dialog models with access to expanded and richer set of persona descriptions. |
| Outcome: | The proposed model outperforms baselines on the Persona-Chat dataset in terms of dialog quality and diversity while achieving persona-consistent and controllable dialog generation. |
Ask what’s missing and what’s useful: Improving Clarification Question Generation using Global Knowledge (2021.naacl-main)
Copied to clipboard
| Challenge: | Existing models that generate clarification questions fail to identify useful information in contexts . human ability to generate fluent and relevant questions is important in reducing ambiguity . |
| Approach: | They propose a model that first identifies what is missing and then generates a question about it. |
| Outcome: | The proposed model outperforms baselines as judged by automatic metrics and humans. |
Latent Factor Models Meets Instructions: Goal-conditioned Latent Factor Discovery without Task Supervision (2025.naacl-long)
Copied to clipboard
Zhouhang Xie, Tushar Khot, Bhavana Dalvi Mishra, Harshit Surana, Julian McAuley, Peter Clark, Bodhisattwa Prasad Majumder
| Challenge: | Instruction-following LLMs have recently allowed systems to discover hidden concepts from a collection of unstructured documents based on a natural language description of the purpose of the discovery (i.e., goal). |
| Approach: | They propose a goal-oriented latent factor discovery system that integrates LLM’s instruction-following ability with statistical models to handle large, noisy datasets where LLM reasoning alone falls short. |
| Outcome: | The proposed system improves task performance by 5-52% over baselines and 1.8 times as often as the best alternative, on average, in human evaluation. |
Evaluating Large Language Models as Generative User Simulators for Conversational Recommendation (2024.naacl-long)
Copied to clipboard
| Challenge: | Large language models show promise in simulating human-like behavior, raising the question of their ability to represent a diverse population of users. |
| Approach: | They propose a protocol to evaluate the degree to which language models can accurately emulate human behavior in conversational recommendation systems. |
| Outcome: | The proposed protocol evaluates five tasks to reveal deviations of language models from human behavior and offers insights on how to reduce deviations with model selection and prompting strategies. |
Achieving Conversational Goals with Unsupervised Post-hoc Knowledge Injection (2022.acl-long)
Copied to clipboard
| Challenge: | Existing neural dialog models lack specificity and informativeness due to limited knowledge available during training. |
| Approach: | They propose a method to extract relevant knowledge from external sources at decoding time and incorporate it into a dialog response. |
| Outcome: | The proposed method in goal-oriented and knowledge-grounded dialog settings shows that human annotators judge the outputs more engaging and informative compared to responses from prior dialog systems. |
Learning Visual-Semantic Embeddings for Reporting Abnormal Findings on Chest X-rays (2020.findings-emnlp)
Copied to clipboard
| Challenge: | Existing work on report generation often trains encoder-decoder networks to generate complete reports, but such models are affected by data bias and face common issues inherent in text generation models. |
| Approach: | They propose a method to identify abnormal findings from radiology images and group them with unsupervised clustering and minimal rules. |
| Outcome: | The proposed method outperforms existing generation models on correctness and text generation metrics. |
Weakly Supervised Named Entity Tagging with Learnable Logical Rules (2021.acl-long)
Copied to clipboard
| Challenge: | Existing methods for building entity tagging systems use weak supervision . previous methods focus on disambiguating entity types based on contexts and expert-provided rules . |
| Approach: | They propose a method that bootstraps high-quality logical rules to train a neural tagger in a fully automated manner. |
| Outcome: | The proposed method outperforms weakly supervised methods on three datasets . it rivals state-of-the-art supervised method with lexicon of over 2,000 terms . |
Justifying Recommendations using Distantly-Labeled Reviews and Fine-Grained Aspects (D19-1)
Copied to clipboard
| Challenge: | Existing approaches to generating reviews struggle to generate justifications that are relevant to users’ decision-making process. |
| Approach: | They propose an ‘extractive’ approach to identify review segments which justify users’ intentions and use it to distantly label massive review corpora and construct large-scale personalized recommendation justification datasets. |
| Outcome: | The proposed model can generate convincing and diverse justifications from massive review corpora and distantly label massive review data. |
SAND: Boosting LLM Agents with Self-Taught Action Deliberation (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large Language Model (LLM) agents finetuned with supervised finetuning may over-commit towards seemingly plausible but suboptimal actions due to limited action space exploration. |
| Approach: | They propose a self-taught actioN deliberation framework that allows LLM agents to explicitly deliberate over candidate actions before committing to one. |
| Outcome: | The proposed framework outperforms state-of-the-art methods on two representative interactive agent tasks and achieves an average 20% improvement over initial finetuning. |
DeCoT: Debiasing Chain-of-Thought for Knowledge-Intensive Tasks in Large Language Models via Causal Intervention (2024.acl-long)
Copied to clipboard
| Challenge: | In large language models, external knowledge is required to augment their internal knowledge through prompts, but this does not guarantee that LLMs can identify and use relevant information in the prompts to conduct chain-of-thought reasoning. |
| Approach: | They propose a structural causal model to formally explain the internal knowledge bias of large language models (LLMs) they review the chain-of-thought (CoT) prompting from a causal perspective and find that biased information from pretrained models can impair LLMs’ reasoning abilities. |
| Outcome: | The proposed model enables more accurate CoT reasoning and enhances LLM generation on knowledge-intensive tasks. |
Evaluating Language Model Pluralism through In-the-wild Crowd Discussions (2026.acl-long)
Copied to clipboard
Gagan Mundada, Rohan Surana, Nandhini Swaminathan, Bodhisattwa Prasad Majumder, Junda Wu, Julian McAuley, Zhouhang Xie
| Challenge: | Existing evaluation methods focus predominantly on multiple-choice and question-answering tasks, leaving open-ended generation largely unaddressed. |
| Approach: | They propose an evaluation framework that assesses LLM pluralism in open-ended generation by comparing outputs against free-form crowd responses. |
| Outcome: | The proposed evaluation framework decomposes ground-truth responses into atomic, non-overlapping claims and evaluates whether LLMs adequately cover this diverse claim space. |
SceneAlign: Aligning Multimodal Reasoning to Scene Graphs in Complex Visual Scenes (2026.acl-long)
Copied to clipboard
Chuhan Wang, Xintong Li, Jennifer Yuntong Zhang, Junda Wu, Chengkai Huang, Lina Yao, Julian McAuley, Jingbo Shang
| Challenge: | Existing preference-based approaches fail to address this challenge by exploiting language priors to bypass visual grounding. |
| Approach: | They propose a framework that leverages scene graphs as structured visual information to perform controllable structural interventions. |
| Outcome: | The proposed framework improves answer accuracy and reasoning faithfulness across seven visual reasoning benchmarks. |
Knowledge-Aware Query Expansion with Large Language Models for Textual and Relational Retrieval (2025.naacl-long)
Copied to clipboard
| Challenge: | Existing methods to generate query expansions focus on enhancing textual similarities between search queries and document corpus, overlooking document relations. |
| Approach: | They propose a knowledge-aware query expansion framework augmenting LLMs with structured document relations from knowledge graph (KG) they leverage document texts as rich KG node representations and use document-based relation filtering for their method. |
| Outcome: | The proposed framework augments LLMs with structured document relations from knowledge graph (KG) Extensive experiments on three datasets of diverse domains show the advantages compared against state-of-the-art methods on textual and relational semi-structured retrieval. |
Generating Personalized Recipes from Historical User Preferences (D19-1)
Copied to clipboard
| Challenge: | Existing methods to recipe generation are unable to create recipes for users with culinary preferences but incomplete knowledge of ingredients in specific dishes. |
| Approach: | They propose to expand a name and incomplete ingredient details into complete natural-text instructions aligned with the user’s historical preferences. |
| Outcome: | The proposed model generates plausible recipes from user-aware representations of recipes from 180K recipes and 700K interactions. |
Mitigating Hallucination in Fictional Character Role-Play (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Influence of parametric knowledge of large language models (LLMs) often causes role-playing characters to act out of character and hallucinate about things outside the scope of their knowledge. |
| Approach: | They propose a method that modulates the influence of parametric knowledge using a pre-calibrated confidence threshold to mitigate hallucination in fictional character role-play. |
| Outcome: | The proposed method reduces the factual accuracy of generated responses by 18% for adversarial questions and 44% in temporal hallucination for time-sensitive interviews. |