Papers by Zaiqiao Meng
Mixture-of-Partitions: Infusing Large Biomedical Knowledge Graphs into BERT (2021.emnlp-main)
Copied to clipboard
| Challenge: | Infusing factual knowledge into pre-trained models is fundamental for many knowledge-intensive tasks. |
| Approach: | They propose an infusion approach that partitions a large knowledge graph into smaller sub-graphs and infuses their specific knowledge into various BERT models using lightweight adapters. |
| Outcome: | The proposed approach improves the underlying BERTs and achieves new SOTA performance on six downstream tasks. |
Time to Revisit Exact Match (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Temporal question answering is an established method for evaluating temporal reasoning in large language models. |
| Approach: | They propose a numerical estimation task where all questions require a numeric, temporal answer, allowing us to evaluate models beyond EM. |
| Outcome: | The proposed model responses are based on a numerical estimation task and are distilled from Test of Time and TempTabQA. |
EvoAgentX: An Automated Framework for Evolving Agentic Workflows (2025.emnlp-demos)
Copied to clipboard
| Challenge: | Existing MAS frameworks often require manual workflow configuration and lack native support for dynamic evolution and performance optimization. |
| Approach: | They propose an open-source platform that automates generation, execution, and evolutionary optimization of multi-agent workflows. |
| Outcome: | The proposed platform automates generation, execution, and evolutionary optimization of multi-agent workflows. |
RadEval: A framework for radiology text evaluation (2025.emnlp-demos)
Copied to clipboard
Justin Xu, Xi Zhang, Javid Abderezaei, Julie Bauml, Roger Boodoo, Fatemeh Haghighi, Ali Ganjizadeh, Eric Brattain, Dave Van Veen, Zaiqiao Meng, David W Eyre, Jean-Benoit Delbrouck
| Challenge: | Evaluating automated radiology report generation systems remains a fundamental challenge in the development of safe, accurate, and clinically useful medical AI. |
| Approach: | They propose a unified, open-source framework for evaluating radiology texts that consolidates a diverse range of metrics from classic ngram overlap (BLEU) and contextual measures (BERTScore) to clinical concept-based scores (GREEN). |
| Outcome: | The framework consolidates a diverse range of metrics from ngram overlap (BLEU) and contextual measures (BERTScore) to clinical concept-based scores (F1CheXbert, F1RadGraph, RaTEScore, SRR-BERT, TemporalEntityF1) and advanced LLMbased evaluators (GREEN). |
CCD: Mitigating Hallucinations in Radiology MLLMs via Clinical Contrastive Decoding (2026.findings-acl)
Copied to clipboard
| Challenge: | Multimodal large language models generate medical hallucinations due to over-sensitivity to clinical sections. |
| Approach: | They propose a framework that integrates structured clinical signals from task-specific radiology expert models. |
| Outcome: | The proposed framework improves overall performance on radiology report generation (RRG) on the MIMIC-CXR dataset, it yields up to 17% improvement in RadGraph-F1. |
Biomedical Named Entity Recognition via Dictionary-based Synonym Generalization (2023.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for biomedical named entity recognition require laborious human effort. |
| Approach: | They propose a Synonym Generalization framework that recognizes biomedical concepts using span-based predictions. |
| Outcome: | The proposed framework outperforms dictionary-based approaches on a wide range of benchmarks. |
TRACE the Evidence: Constructing Knowledge-Grounded Reasoning Chains for Retrieval-Augmented Generation (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Existing retrievers are not perfect and often include irrelevant documents in the retrieved set. |
| Approach: | They propose to construct knowledge-grounded reasoning chains from retrieved documents to integrate supporting evidence into RAG models. |
| Outcome: | The proposed model achieves an average performance improvement of 14.03% on three multi-hop QA datasets. |
Few-Shot Table-to-Text Generation with Prototype Memory (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Neural table-to-text generation models are data-hungry and require large amounts of training data to learn the mapping between tables and texts. |
| Approach: | They propose a framework for table-to-text generation under the few-shot scenario that uses retrieved prototypes and a prototype selector to bridge the structural gap between tables and texts. |
| Outcome: | The proposed framework significantly improves the model performance on three benchmark datasets with state-of-the-art models. |
FusionDTI: Fine-grained Binding Discovery with Token-level Fusion for Drug-Target Interaction (2025.findings-emnlp)
Copied to clipboard
| Challenge: | despite advances in DTI models, models often struggle to capture fine-grained interactions between drugs and proteins. |
| Approach: | They propose a novel drug-target interaction model that uses a token-level module to learn fine-grained information for drug-target interactions. |
| Outcome: | The proposed model learns fine-grained information for drug-target interaction . it mitigates sequence fragment invalidation and incorporates the structure-aware vocabulary of target proteins . |
Can We Edit LLMs for Long-Tail Biomedical Knowledge? (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing knowledge editing methods can enhance LLMs' performance on long-tail biomedical knowledge, but their performance on high-frequency popular knowledge remains inferior to that on high frequency popular knowledge. |
| Approach: | They conduct the first comprehensive study to investigate the effectiveness of knowledge editing methods for editing long-tail biomedical knowledge. |
| Outcome: | The proposed methods improve LLMs' performance on long-tail biomedical knowledge, but their performance on high-frequency popular knowledge remains inferior even after editing. |
Rewire-then-Probe: A Contrastive Recipe for Probing Biomedical Knowledge of Pre-trained Language Models (2022.acl-long)
Copied to clipboard
| Challenge: | Despite the growing progress of probing knowledge for pre-trained language models, specialised areas such as the biomedical domain are vastly under-explored. |
| Approach: | They propose a biomedical knowledge probing benchmark, MedLAMA, constructed based on the Unified Medical Language System (UMLS) Metathesaurus. |
| Outcome: | The proposed approach pushes the acc@10 to 28%, but the performance gap remains notable. |
Libra: Leveraging Temporal Images for Biomedical Radiology Analysis (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing methods for radiology report generation rely on single-image analysis or rule-based heuristics to process multiple images. |
| Approach: | They propose a temporal-aware MLLM tailored for chest X-ray report generation that combines a radiology-specific image encoder with a novel Temporal Alignment Connector. |
| Outcome: | The proposed model sets new standards in clinical relevance and lexical accuracy on the MIMIC-CXR dataset. |
REANO: Optimising Retrieval-Augmented Reader Models through Knowledge Graph Generation (2024.acl-long)
Copied to clipboard
| Challenge: | Existing knowledge graphs suffer from incompleteness and lack information critical for answering given questions. |
| Approach: | They propose to enhance the open domain question answering model with a knowledge graph generation module that generates KGs from the passages and an answer predictor. |
| Outcome: | The proposed model improves the exact match score by 2.7% on the EntityQuestion dataset, with an average improvement of 1.8% across all the datasets. |
To Generate or to Retrieve? On the Effectiveness of Artificial Contexts for Medical Open-Domain Question Answering (2024.acl-long)
Copied to clipboard
| Challenge: | Medical open-domain question answering requires substantial access to specialized knowledge. |
| Approach: | They propose a framework that generates multiple-choice questions from a set of open-book parameters and a small-scale reader that can outcompete closed-book questions by 706x using fewer parameters. |
| Outcome: | The proposed framework outperforms closed-book models on MedQA-USMLE, MedMCQA, and MMLU while using up to 706x fewer parameters. |
SPARKLE: A Structured and Plug-and-play Agentic Retrieval Policy for Adaptive RAG Models (2026.acl-long)
Copied to clipboard
| Challenge: | Existing methods for integrating external knowledge rely on frozen large language models without explicit supervision or require costly LLM finetuning. |
| Approach: | They propose a structured and plug-and-play agentic retrieval policy with an additional proxy model to control the retrieval process. |
| Outcome: | Experiments on three in-domain and four out-of-domain QA benchmarks show that SPARKLE outperforms state-of the-art adaptive RAG models, achieving average improvements of 9.17% and 2.85%, respectively. |
KiRAG: Knowledge-Driven Iterative Retriever for Enhancing Retrieval-Augmented Generation (2025.acl-long)
Copied to clipboard
| Challenge: | Iterative retrieval-augmented generation models are difficult to use for multihop question answering (QA) . their retrieval processes can be disrupted by irrelevant documents or factually inaccurate chain-of-thoughts . |
| Approach: | They propose a knowledge-driven iterative retriever model that decomposes documents into knowledge triples and performs iterativ retrieval with these triples to enable a factually reliable retrieval process. |
| Outcome: | The proposed model outperforms existing iRAG models with an average improvement of 9.40% in R@3 and 5.14% in F1 on multi-hop QA datasets. |
Failure Modes in Multi-Hop QA: The Weakest Link Effect and the Recognition Bottleneck (2026.acl-long)
Copied to clipboard
| Challenge: | Existing studies have identified a position bias in Large Language Models that causes them to overlook information at certain positions. |
| Approach: | They propose a semantic probe to disentangle position bias in Large Language Models . they propose MFAI to steer attention towards selected positions . |
| Outcome: | The proposed model can locate and integrate information at certain positions even in noisy, long-context settings. |
TaCL: Improving BERT Pre-training with Token-aware Contrastive Learning (2022.findings-naacl)
Copied to clipboard
| Challenge: | Existing pre-trained MLMs produce an anisotropic distribution of token representations . this is not ideal for tasks that require discriminative semantic meanings of distinct tokens - a problem that exists in pre-training models . |
| Approach: | They propose a continual pre-training approach that encourages BERT to learn an isotropic distribution of token representations. |
| Outcome: | The proposed approach improves on a wide range of English and Chinese benchmarks. |
Revisiting Parameter-Efficient Tuning: Are We Really There Yet? (2022.emnlp-main)
Copied to clipboard
| Challenge: | Pretrained language models (PLMs) are used as backbones to be combined with additional parameters and finetuned on downstream tasks in an end-to-end manner. |
| Approach: | They propose to use a fraction of parameters to tune pretrained language models (PLMs) this is the first comprehensive investigation into the training and evaluation of PETuning methods. |
| Outcome: | The proposed methods have been validated and tested with a rigorous evaluation protocol and have shown that they are unstable and inconsistent. |
GenKIE: Robust Generative Multimodal Document Key Information Extraction (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Key information extraction (KIE) is a key application for information retrieval and text mining. |
| Approach: | They propose a novel generative end-to-end model, named GenKIE, to address the KIE task. |
| Outcome: | The proposed model generalizes over different types of documents and achieves state-of-the-art results. |
Self-Alignment Pretraining for Biomedical Entity Representations (2021.naacl-main)
Copied to clipboard
| Challenge: | Existing approaches to self-supervised learning of biomedical entities are limited in the biomedic domain. |
| Approach: | They propose a pretraining scheme that self-aligns the representation space of biomedical entities. |
| Outcome: | The proposed framework achieves state-of-the-art on six MEL benchmarking datasets. |
MANNER: A Variational Memory-Augmented Model for Cross Domain Few-Shot Named Entity Recognition (2023.acl-long)
Copied to clipboard
| Challenge: | Named Entity Recognition (NER) is a fundamental NLP task that aims at classifying mention spans into entity types. |
| Approach: | They propose a variational memory-augmented few-shot named entity recognition model that uses a memory module to store information from source domain and retrieve relevant information from the memory to augment few-shot task in target domain. |
| Outcome: | The proposed model can adapt the learned knowledge from source domain to target domain and achieve superior performance on English and Chinese cross domain few-shot NER datasets. |
Can Pretrained Language Models (Yet) Reason Deductively? (2023.eacl-main)
Copied to clipboard
| Challenge: | Acquiring factual knowledge with Pretrained Language Models (PLMs) has attracted increasing attention, showing promising performance in many knowledge-intensive tasks. |
| Approach: | They conduct a comprehensive evaluation of the learnable deductive reasoning capability of pretrained language models and compare their performance against simple adversarial surface form edits. |
| Outcome: | The models are able to generalise learned logic rules and perform inconsistently against simple adversarial surface form edits, but catastrophically forget the previously learnt knowledge. |
Can We Instruct LLMs to Compensate for Position Bias? (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Recent studies reveal that position bias in large language models (LLMs) leads to difficulty in accessing information retrieved from the retriever. |
| Approach: | They propose to direct LLMs to allocate more attention towards a selected segment of the context through prompting. |
| Outcome: | The proposed approach improves the performance of large language models by promoting instruction with an exact document index. |