Papers by Zuchao Li

GKT: A Novel Guidance-Based Knowledge Transfer Framework For Efficient Cloud-edge Collaboration LLM Deployment (2024.findings-acl)

Challenge:	Existing models for event coreference resolution are based on entity-level tasks, but event coreferent resolution is a challenge.
Approach:	They propose a model that learns and integrates multiple representations from event alone and event pair on the basis of event but not entity as before.
Outcome:	The proposed model achieves new state-of-the-art on the ACE 2005 benchmark, demonstrating the effectiveness of the proposed framework.

Copied to clipboard

Hypergraph based Understanding for Document Semantic Entity Recognition (2024.acl-long)

Challenge:	Existing methods of acceleration require fine-tuning of considerably large models, such as Llama-7B, posing a challenge for average users.
Approach:	They propose a Guidance-based Knowledge Transfer framework that leverages a larger LLM as a 'teacher' and a smaller 'student' model to finalize responses.
Outcome:	The proposed framework achieves a maximum accuracy improvement of 14.18%, along with a 10.72 times speed-up on GSM8K and an accuracy improvement 14.00% along with 7.73 times speed up in CSQA.

Copied to clipboard

Qiwei Li, Zuchao Li, Ping Wang, Haojun Ai, Hai Zhao

Memory-R1: Enhancing Large Language Model Agents to Manage and Utilize Memories via Reinforcement Learning (2026.acl-long)

Challenge:	Existing document understanding models focus on entity categories while ignoring the extraction of entity boundaries.
Approach:	They propose a hypergraph attention document semantic entity recognition framework which uses hypergraph focus to focus on entity boundaries and entity categories at the same time.
Outcome:	The proposed framework can improve the performance of existing models on FUNSD, CORD, XFUND and SROIE.

Copied to clipboard

Sikuan Yan, Xiufeng Yang, Zuchao Huang, Ercong Nie, Zifeng Ding, Zonggen Li, Xiaowen Ma, Jinhe Bi, Kristian Kersting, Jeff Z. Pan, Hinrich Schuetze, Volker Tresp, Yunpu Ma

RACER: Retrieval-Augmented Contextual Rapid Speculative Decoding (2026.findings-acl)

Challenge:	Large Language Models (LLMs) are stateless and limited by a finite context window, preventing them from maintaining knowledge across long conversations or evolving tasks.
Approach:	They propose a reinforcement learning framework that empowers LLMs to actively manage external memory through two specialized agents.
Outcome:	The proposed framework outperforms baselines and benchmarks across diverse question types, three benchmarks, and multiple model scales.

Copied to clipboard

Zihong Zhang, Zuchao Li, Lefei Zhang, Ping Wang, Hai Zhao

Information Retrieval

Challenge:	Existing methods for decoding large language models generate one token per step, causing high inference latency.
Approach:	They propose a method that integrates retrieved exact patterns with logit-driven future cues.
Outcome:	Experiments on Spec-Bench, HumanEval, and MGSM-ZH show that RACER outperforms training-free methods and accelerates inference.

TRACE: Traversal Retrieval-Augmented Chain of Evidence for Document Understanding (2026.acl-long)

Copied to clipboard

Liqi He, Zuchao Li, Hao Huang, Ping Wang

VCB Bench: An Evaluation Benchmark for Audio-Grounded Large Language Model Conversational Agents (2026.findings-acl)

Challenge:	Long-context Document Visual Question Answering (DocVQA) methods struggle with visual semantics or handling finite context windows.
Approach:	They propose a new approach to longcontext document visual question answering that transforms retrieval into adaptive evidence chain construction using a Bi-Layered Graph.
Outcome:	The proposed approach achieves an average accuracy improvement of 14.07% on M5BookVQA and exhibits robust generalization with a 13.38% gain across four established benchmarks.

Copied to clipboard

Jiliang Hu, Wenfu Wang, Zuchao Li, Chenxing Li, Yiyang Zhao, Hanzhao Li, Liqiang Zhang, Meng Yu, Dong Yu

Dialogue Systems & Conversational Agents

Challenge:	despite advances in multimodal conversational systems, current benchmarks lack comprehensive evaluation across key dimensions.
Approach:	They propose a Chinese benchmark built exclusively on real human speech to fill this gap . they assess LALMs across three complementary axes: instruction following, knowledge understanding, robustness .
Outcome:	VCB Bench assesses LALMs across three complementary axes: instruction following, knowledge understanding, and robustness . VCBM Bench provides reproducible and fine-grained framework for Chinese voice chat bots . results show significant performance disparities and offer tangible insights for future improvements .

BiBL: AMR Parsing and Generation with Bidirectional Bayesian Learning (2022.coling-1)

Copied to clipboard

Ziming Cheng, Zuchao Li, Hai Zhao

Can Large Language Models Be Good Language Teachers? (2025.emnlp-main)

Challenge:	Existing approaches to AMR focus on one-side improvements despite the duality of the two tasks . instead, we propose data-efficient Bidirectional Bayesian learning (BiBL) to facilitate bidirectional information transition.
Approach:	They propose a data-efficient bidirectional Bayesian learning approach to facilitate bidirectional information transition by adopting a single-stage multitasking strategy.
Outcome:	The proposed model outperforms existing models on benchmark datasets without extra training data.

Copied to clipboard

LiQing Xu, Qiwei Li, Tianshuo Peng, Zuchao Li, Hai Zhao, Ping Wang

Grammatical Error Correction as GAN-like Sequence Labeling (2021.findings-acl)

Challenge:	Large language models (LLMs) have achieved remarkable success across diverse domains, but their potential as effective language teachers remains inadequately assessed.
Approach:	They propose a framework to evaluate Chinese language teachers' pedagogical competence against international standards.
Outcome:	The proposed framework evaluates 13 latest multilingual and Chinese LLMs against international standards for Chinese language teachers.

Copied to clipboard

Kevin Parnow, Zuchao Li, Hai Zhao

ToM: Leveraging Tree-oriented MapReduce for Long-Context Reasoning in Large Language Models (2025.emnlp-main)

Challenge:	Traditional GEC models learn from sentences with fixed error rates . sequence labeling approaches suffer from a couple of key problems .
Approach:	They propose a GAN-like sequence labeling model with a grammatical error detector and a generator to correct grammamatical errors.
Outcome:	The proposed model improves the state-of-the-art in GEC and improves on benchmarks.

Copied to clipboard

Jiani Guo, Zuchao Li, Jie Wu, Qianren Wang, Yun Li, Lefei Zhang, Hai Zhao, Yujiu Yang

Nested Named Entity Recognition as Corpus Aware Holistic Structure Parsing (2022.coling-1)

Challenge:	Experimental results show ToM outperforms existing divide-and-conquer frameworks . RAG relies on similarity-based rankings to retrieve and reason over chunks based on logical coherence .
Approach:	They propose a Tree-oriented MapReduce framework for long-context reasoning . it leverages the hierarchical structure of long documents by constructing a DocTree .
Outcome:	Experimental results show that ToM outperforms existing divide-and-conquer frameworks and RAGs . the proposed framework improves logical coherence and long-context reasoning on 70B+ LLMs compared to existing approaches .

Copied to clipboard

Yifei Yang, Zuchao Li, Hai Zhao

What Works and Doesn’t Work, A Deep Decoder for Neural Machine Translation (2022.findings-acl)

Challenge:	Named entity recognition is a natural language processing task . nested NER is based on a linear structure, but there is no research on applying corpus-level information to NER.
Approach:	They propose a holistic structure parsing algorithm to reveal the entire NEs in a sentence . they introduce points-wise mutual information and other frequency features from corpus-aware statistics .
Outcome:	The proposed model outperforms existing models on widely-used benchmarks and achieves state-of-the-art.

Copied to clipboard

Zuchao Li, Yiran Wang, Masao Utiyama, Eiichiro Sumita, Hai Zhao, Taro Watanabe

From AR to Diffusion: Efficiently Adapting Large Language Models with Strictly Causal and Elastic Horizons (2026.acl-long)

Challenge:	Deep learning has demonstrated performance advantages in a wide range of natural language processing tasks.
Approach:	They propose to deepen the decoder layer in a Transformer model to reduce the difficulty of deep learning.
Outcome:	The proposed method can deepen the model on both the encoder and decoder at the same time, resulting in a deeper model and improved performance.

Copied to clipboard

Xiangyu Ma, Teng Xiao, Zuchao Li, Lefei Zhang

Faster In-Context Learning for LLMs via N-Gram Trie Speculative Decoding (2025.emnlp-main)

Challenge:	Autoregressive (AR) models rely on bidirectional attention, creating a structural mismatch with pre-trained Autoregression models.
Approach:	They propose a framework that efficiently adapts autoregressive (AR) models to the diffusion paradigm.
Outcome:	The proposed framework reduces training costs by orders of magnitude while maintaining state-of-the-art performance.

Copied to clipboard

Jinglin Chen, Qiwei Li, Zuchao Li, Baoyuan Qi, Liu Guoming, Haojun Ai, Hai Zhao, Ping Wang

Syntax-aware Multilingual Semantic Role Labeling (D19-1)

Challenge:	In-Context Learning (ICL) is a key method in prompt engineering, but its long retrieved contexts and limited token throughput will slow reasoning speeds.
Approach:	They propose a method that leverages the overlap between context and model output to generate drafts from the context.
Outcome:	The proposed method achieves the highest mean speedup on Vicuna-7B, Llama2-7B-Chat, and Llma3-8B-Instruct tasks.

Copied to clipboard

Shexia He, Zuchao Li, Hai Zhao

Vista-LLM: Decoupled Query-Guided Visual Token Pruning for Efficient Long-Video Large Language Models (2026.acl-long)

Challenge:	Existing work on semantic role labeling (SRL) on English has focused on syntactic integration and enhanced word representation.
Approach:	They propose a method guided by syntactic rule to prune arguments to integrate syntax into multilingual SRL model simply and effectively.
Outcome:	The proposed model achieves state-of-the-art results on the CoNLL-2009 benchmarks of all seven languages.

Copied to clipboard

Zhenyu Li, Zuchao Li, Ping Wang, Lefei Zhang, Haojun Ai

IAM: Efficient Inference through Attention Mapping between Different-scale LLMs (2025.acl-long)

Challenge:	Long-video understanding is bottlenecked by the high cost of processing massive visual tokens.
Approach:	They propose a decoupled framework for query-guided visual token pruning . their method reduces visual tokens by 90% and accelerates inference by 98% .
Outcome:	The proposed framework reduces visual tokens by 90% and accelerates inference while retaining over 98% of baseline performance on average.

Copied to clipboard

Yi Zhao, Zuchao Li, Hai Zhao

Green & Sustainable NLP

Challenge:	Large language models (LLMs) are a challenge due to their internal reasoning processes.
Approach:	They propose an algorithm that can optimize attention matrices by performing attention mapping between small and large LLMs.
Outcome:	The proposed framework can reduce KV cache usage by 22.1% and accelerate prefill by 15% without sacrificing performance.

Restricted or Not: A General Training Framework for Neural Machine Translation (2022.acl-srw)

Copied to clipboard

Zuchao Li, Masao Utiyama, Eiichiro Sumita, Hai Zhao

Machine Translation

Challenge:	Existing work imposes constraints on beam search decoding, which limits the concurrent processing ability of the model in deployment.
Approach:	They propose a general training framework that allows a model to support both restricted and unrestricted translations by adopting an additional auxiliary training process without constraining the decoding process.
Outcome:	The proposed training framework is tested on simulated and original benchmarks.

Reference Language based Unsupervised Neural Machine Translation (2020.findings-emnlp)

Copied to clipboard

Zuchao Li, Hai Zhao, Rui Wang, Masao Utiyama, Eiichiro Sumita

Text Generation

Challenge:	Existing approaches to use a common language as an auxiliary for better translation have a long tradition in machine translation.
Approach:	They propose a reference language-based framework for unsupervised neural machine translation that uses only one auxiliary language as an auxiliary for better translation.
Outcome:	The proposed framework improves the quality of pivot translation over a baseline that uses only one auxiliary language.

XQuant: Achieving Ultra-Low Bit KV Cache Quantization with Cross-Layer Compression (2025.emnlp-main)

Copied to clipboard

Haoqi Yang, Yao Yao, Zuchao Li, Baoyuan Qi, Liu Guoming, Hai Zhao

Challenge:	Large Language Models (LLMs) have demonstrated remarkable capabilities across diverse tasks. however, their extensive memory requirements present significant challenges for deployment in resource-constrained environments.
Approach:	They propose a training-free framework that achieves ultra-low equivalent bit-width KV cache quantization.
Outcome:	The proposed framework outperforms state-of-the-art methods on TruthfulQA and LongBench.

Reorder and then Parse, Fast and Accurate Discontinuous Constituency Parsing (2022.emnlp-main)

Copied to clipboard

Kailai Sun, Zuchao Li, Hai Zhao

A Unified Syntax-aware Framework for Semantic Role Labeling (D18-1)

Challenge:	Discontinuous constituency parsing is still being developed for its efficiency and accuracy are far behind its continuous counterparts.
Approach:	They propose to transform a discontinuous constituent tree into a pseudo-continuous one by reordering words in the sentence.
Outcome:	The proposed method can transform a discontinuous constituent tree into a pseudo-continuous one by parsing and performing actions on three classical discontinuous constituency treebanks.

Copied to clipboard

Zuchao Li, Shexia He, Jiaxun Cai, Zhuosheng Zhang, Hai Zhao, Gongshen Liu, Linlin Li, Luo Si

Moon IME: Neural-based Chinese Pinyin Aided Input Method with Customizable Association (P18-4)

Challenge:	Syntactic information has been paid a great attention over the role of enhancing SRL . but the gap between syntax-aware and syntax-gnostic SRL is smaller . a new framework proposes syntax-based SRL for a wide range of NLP tasks .
Approach:	They propose to extend existing models to investigate more effective ways of incorporating syntax into sequential neural networks.
Outcome:	The proposed framework outperforms existing models on CoNLL-2009 benchmarks in English and Chinese.

Copied to clipboard

Yafang Huang, Zuchao Li, Zhuosheng Zhang, Hai Zhao

Machine Translation

Challenge:	a pinyin input method engine (IME) allows users to input Chinese into a computer by typing pinyan through the common keyboard.
Approach:	They present a pinyin IME that integrates neural machine translation and IR to offer amusive and customizable association ability.
Outcome:	The Moon IME integrates neural machine translation and IR to offer amusive association ability.

DAC: A Dynamic Attention-aware Approach for Task-Agnostic Prompt Compression (2025.acl-long)

Copied to clipboard

Yi Zhao, Zuchao Li, Hai Zhao, Baoyuan Qi, Liu Guoming

MiSS: An Assistant for Multi-Style Simultaneous Translation (2021.emnlp-demo)

Challenge:	Existing methods rely on information entropy as the metric to compress lexical units, but ignore attention-critical tokens and information . recent advent of In-Context Learning (ICL), Chain-of-Thought (CoT), and Retrieval Augmented Generation (RAG) technologies has significantly invigorated the landscape of applications based on Large Language Models (LLMs).
Approach:	They propose a dynamic attention-aware approach to task-agnostic prompt compression . they integrate entropy and attention information to achieve fine-grained prompt compression.
Outcome:	Experiments show that the proposed approach improves across tasks and LLMs.

Copied to clipboard

Zuchao Li, Kevin Parnow, Masao Utiyama, Eiichiro Sumita, Hai Zhao

Machine Translation

Challenge:	MiSS is a multi-style simultaneous translation assistant . it has five key features: high translation accuracy, simultaneous translation, flexibility, and measurable translation quality.
Approach:	They propose an assistant system for multi-style simultaneous translation that provides a complete translation experience for machine translation users.
Outcome:	The proposed system improves translation efficiency and performance by combining machine translation, grammatical error correction, and interactive edits.

NOTA: Multimodal Music Notation Understanding for Visual Large Language Model (2025.findings-naacl)

Copied to clipboard

Mingni Tang, Jiajia Li, Lu Yang, Zhiqiang Zhang, Jinhao Tian, Zuchao Li, Lefei Zhang, Ping Wang

Faster MoE LLM Inference for Extremely Large Models (2026.findings-acl)

Challenge:	Existing general-domain visual language models lack ability of music notation understanding . Symbolic music is represented in two distinct forms: auditory music and symbolic music .
Approach:	They propose to train a multimodal music notation model using a large-scale dataset . they use cross-modal alignment to train the model for music notations analysis .
Outcome:	The proposed model improves on music understanding by training with a multimodal music notation model.

Copied to clipboard

Haoqi Yang, Luohe Shi, Qiwei Li, Zuchao Li, Ping Wang, Hao Huang, Hai Zhao

Responsible & Trustworthy NLP

Challenge:	Existing inference optimizations for coarse-grained Mixture-of-Experts models implicitly assume a fixed activation budget, which is poorly understood.
Approach:	They propose a training-free policy that adapts token-level activation using router confidence and entropy while remaining within the model’s original budget.
Outcome:	The proposed skipping policy can provide substantial throughput gains, but optimal static schedules vary significantly across models and routing mechanisms.

From Parameters to Performance: A Data-Driven Study on LLM Structure and Development (2025.emnlp-main)

Copied to clipboard

Suqing Wang, Zuchao Li, Shi Luohe, Bo Du, Hai Zhao, Yun Li, Qianren Wang

Challenge:	Large language models have revolutionized a wide range of domains, driving significant advancements in both technology and real-world applications.
Approach:	They present a large-scale dataset encompassing diverse open-source LLM structures and their performance across multiple benchmarks.
Outcome:	The proposed model validates the relationship between structural configurations and performance across multiple benchmarks and further corroborates the findings using mechanistic interpretability techniques.

BoYaEval: Evaluating Multimodal Large Language Models on Understanding Ancient Chinese Musical Scores (2026.acl-long)

Copied to clipboard

Jiajia Li, Weizhi Xue, Yao Yao, Qiwei Li, null Chenchong, Zuchao Li, Ping Wang, Hai Zhao

Zuchao Li, Jiaxun Cai, Shexia He, Hai Zhao

Challenge:	Multimodal Large Language Models excel in general tasks but struggle with specialized, structured cultural symbols.
Approach:	They evaluate 21 leading MLLMs and compare their performance to a benchmark for Ancient Chinese musical notation.
Outcome:	The benchmark evaluates 21 leading MLLMs on five types of ancient Chinese music notation systems.

Seq2seq Dependency Parsing (C18-1)

Copied to clipboard

Contextualized Semantic Distance between Highly Overlapped Texts (2023.findings-acl)

Challenge:	Recent trend for dependency parsing is adopting neural networks due to their significant success in a wide range of applications.
Approach:	They propose a sequence to sequence (seq2seque) dependency parser that predicts the relative position of head for each word.
Outcome:	The proposed parser achieves 94.11% UAS on PTB and 88.78% UAS .

Copied to clipboard

Letian Peng, Zuchao Li, Hai Zhao

Teaching Your Models to Understand Code via Focal Preference Alignment (2025.emnlp-main)

Challenge:	Conventional semantic metrics are based on word representations and are vulnerable to disturbance of overlapped components with similar representations.
Approach:	They propose a mask-and-predict strategy to evaluate the semantic distance between the overlapped sentences using words in the longest common sequence as neighboring words and use masked language modeling to predict their positions.
Outcome:	The proposed method outperforms the state-of-the-art in domain adaption by a huge margin.

Copied to clipboard

Jie Wu, Haoling Li, Xin Zhang, Xiao Liu, Yangyu Huang, Jianwen Luo, Yizhen Zhang, Zuchao Li, Ruihang Chu, Yujiu Yang, Scarlett Li

Programming Languages in NLP

Challenge:	Existing methods for supervised fine-tuning focus on unit test feedback to construct preference pairs.
Approach:	They propose a preference alignment framework that mimics human iterative debugging to refine Code LLMs.
Outcome:	Experiments show that Preference Learning improves on BigCodeBench and BigCodeBind tasks.

TrigReason: Trigger-Based Collaboration between Small and Large Reasoning Models (2026.findings-acl)

Copied to clipboard

Yi Zhao, Yajuan Peng, Cam-Tu Nguyen, Zuchao Li, Xiaoliang Wang, Xiaoming Fu, Hai Zhao

Reasoning

Challenge:	Large Reasoning Models suffer from high inference latency due to autoregressive reasoning . SpecReason adopts a polling-based design that repeatedly invokes the LRM for verification at every step .
Approach:	They propose a trigger-based collaborative reasoning framework that delegates most reasoning to the SRM and activates LRM intervention only when necessary.
Outcome:	The proposed framework reduces latency and API cost by 73.3% under edge–cloud conditions.

SpindleKV: A Novel KV Cache Reduction Method Balancing Both Shallow and Deep Layers (2025.acl-long)

Copied to clipboard

Zicong Tang, Shi Luohe, Zuchao Li, Baoyuan Qi, Liu Guoming, Lefei Zhang, Ping Wang

Challenge:	Large language models (LLMs) have impressive capabilities across various fields, but their widespread use is facing a severe and realistic challenge, which is their high demand for GPU memory.
Approach:	They propose a KV cache reduction method which balances both shallow and deep layers by using an attention weight based eviction method and a codebook based replacement approach.
Outcome:	The proposed method reduces the KV cache for shallower layers while preserving similar or even better model performance.

The Music Maestro or The Musically Challenged, A Massive Music Evaluation Benchmark for Large Language Models (2024.findings-acl)

Copied to clipboard

Jiajia Li, Lu Yang, Mingni Tang, Chenchong Chenchong, Zuchao Li, Ping Wang, Hai Zhao

Seeking Common but Distinguishing Difference, A Joint Aspect-based Sentiment Analysis Model (2021.emnlp-main)

Challenge:	Existing benchmarks to evaluate LLMs' capabilities are inadequate for assessing their musical capabilities.
Approach:	They propose to use a large-scale music benchmark specifically designed to evaluate the music-related capabilities of large language models (LLMs).
Outcome:	The proposed framework evaluates 16 large language models in the domain of music.

Copied to clipboard

Hongjiang Jing, Zuchao Li, Hai Zhao, Shu Jiang

FSUIE: A Novel Fuzzy Span Mechanism for Universal Information Extraction (2023.acl-long)

Challenge:	Existing models focus on aspect term extraction, opinion term extraction and sentiment polarity classification but ignore the difference.
Approach:	They propose a joint aspect-based sentiment analysis task that focuses on the difference between the two tasks to improve the model's robustness.
Outcome:	Empirical results show that the proposed model outperforms the previous state-of-the-art on four benchmark datasets.

Copied to clipboard

Tianshuo Peng, Zuchao Li, Lefei Zhang, Bo Du, Hai Zhao

SirLLM: Streaming Infinite Retentive LLM (2024.acl-long)

Challenge:	Existing Universal Information Extraction models rely heavily on span boundaries in data during training, which does not reflect the reality of span annotation challenges.
Approach:	They propose a framework that uses fuzzy spans to model various IE tasks . they propose generative Universal Information Extraction (UIE) to unify various ie tasks based on fuzzy span boundaries .
Outcome:	The proposed framework improves on a series of main IE tasks with small amounts of data and training epochs.

Copied to clipboard

Soft-Prompting with Graph-of-Thought for Multi-modal Representation Learning (2024.lrec-main)

Challenge:	Large Language Models (LLMs) are becoming increasingly prevalent in various domains, requiring a one-off input of overly long texts to maintain a degree of memory.
Approach:	They propose a Streaming Infinite Retentive LLM which allows LLMs to maintain longer memory during infinite-length dialogues without fine-tuning.
Outcome:	The proposed model can achieve stable and significant improvements across different LLMs and tasks, compellingly proving its effectiveness.

Copied to clipboard

Jun Cheng Yang, Zuchao Li, Shuai Xie, Wei Yu, Shijun Li, Bo Du

Syntax for Semantic Role Labeling, To Be, Or Not To Be (P18-1)

Challenge:	Existing approaches to learn multi-modal tasks are based on chain-of-thought . however, human thought processes are non-linear and employ dynamic adjustment and updating mechanisms.
Approach:	They propose a chain-of-thought technique that adjusts the length of the chain to improve the performance of generated prompts.
Outcome:	The proposed model improves multi-modal representation learning in visual, visual, and audio-visual tasks and also has good domain generalization performance due to better reasoning.

Copied to clipboard

Shexia He, Zuchao Li, Hai Zhao, Hongxiao Bai

Unsupervised Neural Machine Translation with Universal Grammar (2021.emnlp-main)

Challenge:	Existing neural SRL models lack syntactic backbone for performance, limiting its use in deep learning.
Approach:	They propose an enhanced argument labeling model with extended korder argument pruning algorithm for effectively exploiting syntactic information.
Outcome:	The proposed model achieves state-of-the-art on the CoNLL-2008 and 2009 benchmarks in English and Chinese.

Copied to clipboard

Zuchao Li, Masao Utiyama, Eiichiro Sumita, Hai Zhao

Multilinguality

Challenge:	Unsupervised machine translation relies on parallel corpora for training, but performance still lags behind traditional supervised machine translators.
Approach:	They propose to leverage shared grammar clues to provide more explicit language parallel signals to enhance the training of unsupervised machine translation models.
Outcome:	The proposed models improve on a common language pair training task in English and german, and use embedding alignments and pretrained language models to synthesize pseudo parallel corpora.

Selective Prefix Tuning for Pre-trained Language Models (2024.findings-acl)

Copied to clipboard

Hongyi Zhang, Zuchao Li, Ping Wang, Hai Zhao

VHASR: A Multimodal Speech Recognition System With Vision Hotwords (2024.emnlp-main)

Challenge:	Existing methods for fine-tuning pre-trained models are time-consuming and memory-inefficient.
Approach:	They propose a method that inserts learnable vectors into each Transformer layer . they propose SL to encourage diversity in prefix tokens .
Outcome:	Extensive experiments validate the effectiveness of Prefix Tuning in sentence and token classification tasks.

Copied to clipboard

Jiliang Hu, Zuchao Li, Ping Wang, Haojun Ai, Lefei Zhang, Hai Zhao

High-order Semantic Role Labeling (2020.findings-emnlp)

Challenge:	Existing models that incorporate audio-related image information do not improve speech recognition performance.
Approach:	They propose a novel approach utilizing audio-related image information and set up a multimodal speech recognition system that uses vision as hotwords to enhance the model’s speech recognition capability.
Outcome:	The proposed model outperforms unimodal ASR model and achieves SOTA among existing image-based multimodal ASL models.

Copied to clipboard

Zuchao Li, Hai Zhao, Rui Wang, Kevin Parnow

A Full End-to-End Semantic Role Labeler, Syntactic-agnostic Over Syntactic-aware? (C18-1)

Challenge:	Experimental results show that high-order structural learning techniques are beneficial to SRL models . high-level features and structure learning are not common in deep neural networks .
Approach:	They propose a high-order graph structure for a neural semantic role labeling model . it explicitly considers the isolated predicate-argument pairs and interaction between them .
Outcome:	The proposed model can explicitly consider the isolated predicate-argument pairs and the interaction between the predicates-argoments pairs.

Copied to clipboard

Jiaxun Cai, Shexia He, Zuchao Li, Hai Zhao

GoT: Effective Graph-of-Thought Reasoning in Language Models (2024.findings-naacl)

Challenge:	Existing models for semantic role labeling are syntax-agnostic, but outperform them on benchmarks.
Approach:	They propose an end-to-end neural model which tackles the SRL problem in one shot . they augment the encoder with a non-linear transformation to distinguish the predicate and the argument .
Outcome:	The proposed model outperforms state-of-the-art syntax-aware SRL systems on CoNLL-2008 and 2009 benchmarks for English and Chinese.

Copied to clipboard

GoT-R1: Internalizing Graph-of-Thought via Structural Reinforcement for High-Density Reasoning (2026.findings-acl)

Challenge:	Recent advances in Large Language Models (LLMs) have been advancing at an unprecedented pace.
Approach:	They propose a graph-based approach which models human thought processes as a chain and as 'graphs' by representing thought units as nodes and connections between them as edges, they capture the non-sequential nature of human thinking and allows for a more realistic modeling of thought processes.
Outcome:	The proposed model improves on a text-only reasoning task and a multimodal reasoning task.

Copied to clipboard

Zuchao Li, Qiwei Li, Yao Yao, Hai Zhao, Lefei Zhang, Bo Du

Reasoning

Challenge:	Chain-of-Thought reasoning suffers from an inherent mechanism flaw: linearity induces overthinking . emergence of Large Language Models (LLMs) has fundamentally redefined artificial intelligence .
Approach:	They propose a framework that replaces verbose linear trajectories with high-density reasoning graphs.
Outcome:	The proposed framework outperforms state-of-the-art models with reduced token overhead.

Enhancing Ancient Chinese Understanding with Derived Noisy Syntax Trees (2023.acl-srw)

Copied to clipboard

Ping Wang, Shitou Zhang, Zuchao Li, Jingrui Hou

Responsible & Trustworthy NLP

Challenge:	Syntactic information is not used in modern Chinese understanding tasks due to the lack of syntactical annotation.
Approach:	They propose a confidence-based syntax encoding network to alleviate the side effects of unsupervised syntax derivation and the incompatibility between ancient and modern Chinese.
Outcome:	The proposed component alleviates side effects from unsupervised syntax derivation and incompatibility between ancient and modern Chinese.

Segment First or Comprehend First? Explore the Limit of Unsupervised Word Segmentation with Large Language Models (2025.acl-long)

Copied to clipboard

Zihong Zhang, Liqi He, Zuchao Li, Lefei Zhang, Hai Zhao, Bo Du

Low-Resource NLP

Challenge:	Existing approaches to measure word segmentation only assess the language model's understanding of the overall meaning of sentences, lacking an evaluation of the language models' understanding capabilities at a fine-grained level.
Approach:	They propose a framework to explore the limit of unsupervised word segmentation with Large Language Models (LLMs) they employ current mainstream LLMs to perform word segmentations across multiple languages .
Outcome:	The proposed method improves on existing methods and combines the advanced pattern recognition capabilities of Aho-Corasick automata with the deep insights of well-pretrained LLMs.

Parsing All: Syntax and Semantics, Dependencies and Spans (2020.findings-emnlp)

Copied to clipboard

Junru Zhou, Zuchao Li, Hai Zhao

Label Drop for Multi-Aspect Relation Modeling in Universal Information Extraction (2025.naacl-long)

Challenge:	Syntactic and semantic structures are key linguistic contextual clues, but few studies have explored how they can be used to improve syntactical parsing.
Approach:	They propose a syntactic and semantic parsing model which integrates syntaktic information in the encoder of neural network and benefits from two representation formalisms in a uniform way.
Outcome:	The proposed model achieves state-of-the-art or competitive results on both span and dependency representations and on Penn Treebank.

Copied to clipboard

Lu Yang, Jiajia Li, En Ci, Lefei Zhang, Zuchao Li, Ping Wang

Dialogue-RAG: Enhancing Retrieval for LLMs via Node-Linking Utterance Rewriting (2025.acl-long)

Challenge:	Extractive UIEs can solve model explosion problems using a relatively small model . single-target instruction UIE enables the extraction of only one type of relation at a time .
Approach:	They propose a model that assigns different relations to different levels for understanding and decision-making.
Outcome:	Experiments show that LDNet outperforms state-of-the-art systems on 9 tasks, 33 datasets . LDnet outperformed state- of-the art systems on single-modal and multi-modal tasks .

Copied to clipboard

Qiwei Li, Teng Xiao, Zuchao Li, Ping Wang, Mengjia Shen, Hai Zhao

Paraphrasing

Challenge:	Large Language Models (LLMs) and Retrieval Augmented Generation (RAG) methods have demonstrated significant potential on tasks across multiple domains.
Approach:	They propose a lightweight IUR model for query rewriting to complete key information in dialogue to enhance retrieval.
Outcome:	The proposed model improves retrieval and generation ability of RAG system in multi-round dialogue scenarios.

CoViPAL: Layer-wise Contextualized Visual Token Pruning for Large Vision-Language Models (2025.findings-emnlp)

Copied to clipboard

Zicong Tang, Ziyang Ma, Suqing Wang, Zuchao Li, Lefei Zhang, Hai Zhao, Yun Li, Qianren Wang

Visual Data in NLP

Challenge:	Existing methods to prune redundant vision tokens struggle in shallow layers due to the lack of contextual information.
Approach:	They propose a layer-wise contextualized visual token pruning method that uses a plug-and-play Pruning Module to prune redundant vision tokens.
Outcome:	The proposed method outperforms training-free pruning methods under equal token budgets and surpasses training based methods with comparable supervision.

KV-Latent: Dimensional-level KV Cache Reduction with Frequency-aware Rotary Positional Embedding (2025.acl-long)

Copied to clipboard

Shi Luohe, Zuchao Li, Lefei Zhang, Baoyuan Qi, Liu Guoming, Hai Zhao

PAR: Training-Free Positional Perturbation and Attention Recycling for Faithful OCR (2026.acl-long)

Challenge:	Large language models (LLMs) based on Transformer Decoders have become the preferred choice for conversational generative AI.
Approach:	They propose a paradigm called KV-Latent to reduce the KV cache footprint and improve inference speed by down-sampling the Key-Value vector dimensions into a latent space.
Outcome:	The proposed paradigm reduces the KV Cache footprint and improves inference speed with a small amount of extra training, less than 1% of pre-training takes.

Copied to clipboard

Yao Yao, Manwen Liao, Weitian Zhang, Zuchao Li, Hai Zhao