Papers by Zhenglu Yang

20 papers
Regret-Now: A Physics-Inspired Regret Framework for Temporal Knowledge Graph Question Answering with LLMs (2026.findings-acl)

Copied to clipboard

Challenge: Large Language Models have impressive results in general reasoning tasks, but they still exhibit a lack of dynamic error-correction.
Approach: They propose a temporal reasoning framework that uses the principle of minimum potential energy to model the reasoning process as a dynamic trajectory moving toward a more stable state.
Outcome: The proposed framework shows consistent gains over strong baselines on two standard TKGQA benchmarks.
Exploring Union and Intersection of Visual Regions for Generating Questions, Answers, and Distractors (2024.emnlp-main)

Copied to clipboard

Challenge: Existing efforts to generate image-related questions, correct answers, or challenge distractors are limited.
Approach: They propose to put the spotlight on different image regions to diversify QADs . they propose a framework that generates each QAD based on a recurrent multimodal encoder .
Outcome: The proposed framework puts the spotlight on different image regions to diversify QADs.
GMH: A General Multi-hop Reasoning Model for KG Completion (2021.emnlp-main)

Copied to clipboard

Challenge: Knowledge graphs are incomplete with many facts missing, causing performance bottlenecks in many applications.
Approach: They propose a general multi-hop reasoning task that can be formulated as a search process and can be extended to long-distance reasoning scenarios.
Outcome: The proposed model improves on baselines in short and long distance reasoning scenarios.
Modeling Temporal-Modal Entity Graph for Procedural Multimodal Machine Comprehension (2022.acl-long)

Copied to clipboard

Challenge: Procedural Multimodal Documents organize textual instructions and corresponding images step by step.
Approach: They propose a novel temporal-modal entity Graph for comprehending PMDs . they propose encoding and reasoning modules to capture textual and visual entities .
Outcome: The proposed model can capture textual and visual entities and trace their temporal-modal evolution.
Curriculum Pre-training for End-to-End Speech Translation (2020.acl-main)

Copied to clipboard

Challenge: End-to-end speech translation requires a powerful encoder to transcribe, understand and learn cross-lingual semantics simultaneously.
Approach: They propose a curriculum pre-training method that includes an elementary course for transcription learning and two advanced courses for understanding the utterance and mapping words in two languages.
Outcome: The proposed method improves on En-De and En-Fr speech translation benchmarks.
Fact-Tree Reasoning for N-ary Question Answering over Knowledge Graphs (2022.findings-acl)

Copied to clipboard

Challenge: Current Question Answering over Knowledge Graphs (KGQA) tasks focus on binary facts, but neglect n-ary facts.
Approach: They propose a new fact-tree reasoning framework that transforms the question into a fact tree and performs iterative fact reasoning on the fact tree to infer the correct answer.
Outcome: The proposed framework performs iterative fact reasoning on the fact tree to infer the correct answer.
MSIA: Adaptive Medical Multimodal Multi-turn Semantic Jailbreak (2026.findings-acl)

Copied to clipboard

Challenge: Medical large language models exhibit high domain specificity and condensed semantics, making them vulnerable to diagnostic errors in real-world clinical settings.
Approach: They propose a framework for modeling and inducing multi-turn medical semantic jailbreaks in clinical dialogues.
Outcome: Experiments on chest X-ray-based multimodal medical dialogues show that MSIA outperforms existing jailbreak methods with an average success rate of 76.67%.
RepSum: Unsupervised Dialogue Summarization based on Replacement Strategy (2021.acl-long)

Copied to clipboard

Challenge: Existing methods to learn vital information from dialogue context with limited data are limited due to limited words in utterances and huge gap between dialogue and its summary.
Approach: They propose an unsupervised strategy to learn vital information from dialogue context . the proposed model uses a hypothetical foundation that a superior summary approximates a replacement of the original dialogue .
Outcome: The proposed model outperforms existing models on a number of datasets.
ACROSS: An Alignment-based Framework for Low-Resource Many-to-One Cross-Lingual Summarization (2023.findings-acl)

Copied to clipboard

Challenge: Existing studies ignore data imbalance in multilingual settings and do not utilize monolingual data.
Approach: They propose a cross-lingual summarization model that aligns cross-linguistic data with high-resource monolingual data via contrastive and consistency loss.
Outcome: The proposed model outperforms baseline models and consistently dominates on 45 language pairs.
Generating Questions, Answers, and Distractors for Videos: Exploring Semantic Uncertainty of Object Motions (2025.findings-acl)

Copied to clipboard

Challenge: Existing video QADs are generated using video captions, incurring significant costs and resulting in bias.
Approach: They propose to use temporal motion to describe video objects to generate diverse QADs focusing on different objects and interactions.
Outcome: The proposed approach improves consistency and diversity of generated QADs on the NExT-QA and Perception Test benchmarks.
Multi-Party Empathetic Dialogue Generation: A New Task for Dialog Systems (2022.acl-long)

Copied to clipboard

Challenge: Existing work on empathetic dialogues focused on the two-party scenario, but multi-party dialogues are pervasive in reality.
Approach: They propose a multi-party empathetic dialogue generation task that uses a static-dynamic model to explore emotion and sensibility.
Outcome: The proposed task is based on a model with static sensibility and dynamic emotion . it achieves state-of-the-art performance in multi-party empathetic dialogue learning .
HyperPELT: Unified Parameter-Efficient Language Model Tuning for Both Language and Vision-and-Language Tasks (2023.findings-acl)

Copied to clipboard

Challenge: Pretraining and fine-tuning are the dominant paradigms in natural language processing.
Approach: They propose a parameter-efficient multitask learning framework that takes trainable hyper-embeddings and visual modality as input and outputs weights for different modules in a pretrained language model.
Outcome: The proposed framework adds fewer trainable parameters in multi-task learning while achieving superior performances and transfer ability compared to state-of-the-art methods.
Listening to Patients: Detecting and Mitigating Patient Misreport in Medical Dialogue System (2025.findings-acl)

Copied to clipboard

Challenge: Medical Dialogue Systems (MDSs) aim to provide automated healthcare support through natural language interactions between patients and system agents.
Approach: They propose a framework that detects misreports and mitigates them by generating controlled clarifying questions.
Outcome: The proposed framework can detect misreports and mitigate them through generating controlled clarifying questions.
M-TRACE: Detecting and Mitigating Time-Anchor Drift via Step-wise Conflict Checking in Temporal Reasoning (2026.findings-acl)

Copied to clipboard

Challenge: Experimental results show that M-TRACE effectively reduces time-anchor drift . external knowledge may be inaccurate while internal knowledge can become outdated .
Approach: They propose a multi-agent reasoning framework for temporal knowledge conflicts . they propose 'TimeConfQA' which guides conflict-aware final reasoning .
Outcome: Experimental results show that M-TRACE reduces time-anchor drift and improves performance on complex temporal question answering tasks.
Well Begun is Half Done: Generator-agnostic Knowledge Pre-Selection for Knowledge-Grounded Dialogue (2023.emnlp-main)

Copied to clipboard

Challenge: Existing knowledge selection methods are costly to learn and difficult to interpret when errors arise in the generated responses.
Approach: They propose a generator-agnostic knowledge selection method to select context-related knowledge among different knowledge structures and variable knowledge requirements.
Outcome: The proposed method can select knowledge accurately in advance and reduce learning, adjustment, and interpretation burden of later models.
Attention Optimization for Abstractive Document Summarization (D19-1)

Copied to clipboard

Challenge: Abstractive summarization models require attention to reproduce the most salient information.
Approach: They propose to use local and global variances to augment the vanilla attention model to reproduce the most salient information and avoid repetitions.
Outcome: The proposed attention refinement unit can reproduce the most salient information and avoid repetitions on CNN/Daily Mail dataset.
JTAV: Jointly Learning Social Media Content Representation by Fusing Textual, Acoustic, and Visual Features (C18-1)

Copied to clipboard

Challenge: Existing studies on learning social media content focus on single modal or bi-modal learning, but this approach is non-trivial and challenging because content is multi-modal and involves several types of data, including text, audio, and image.
Approach: They propose to combine textual, acoustic, and visual information to learn social media content by fusing them jointly.
Outcome: The proposed model outperforms the state-of-the-art approaches on real-world datasets by a large margin.
MM-AVS: A Full-Scale Dataset for Multi-modal Summarization (2021.naacl-main)

Copied to clipboard

Challenge: Multimodal summarization materials lacking a holistic organization by integrating resources from various modalities.
Approach: They propose a multimodal article and video summarization dataset that integrates resources from different modalities.
Outcome: The proposed dataset validates the important assistance role of external information for multimodal summarization.
A Multi-Attention based Neural Network with External Knowledge for Story Ending Predicting Task (C18-1)

Copied to clipboard

Challenge: Existing studies on the topic of common sense story understanding focus on generating guesses for a missing event or concentrating on unsupervised learning.
Approach: They propose to extend attention-based neural network with external knowledge resources to understand temporal stories and predict their endings.
Outcome: The proposed model outperforms state-of-the-art models and external knowledge resources.
Can We Learn Question, Answer, and Distractors All from an Image? A New Task for Multiple-choice Visual Question Answering (2024.lrec-main)

Copied to clipboard

Challenge: Existing studies focus on generating QADs from image and question, but a novel task is needed to generate meaningful questions, correct answers, and challenging distractors.
Approach: They propose a task to generate QADs from images and encode images together . they use contrastive learning to ensure consistency of QAD generated and tested .
Outcome: Empirical evaluations on the benchmark dataset validate the performance of the proposed task.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations