Papers by Peng Shi

Semantics of the Unwritten: The Effect of End of Paragraph and Sequence Tokens on Text Generation with GPT2 (2021.acl-srw)

Challenge:	In this paper, we address the problem of data scarcity for the Hong Kong Cantonese language . due to the popularization of deep learning, ASR technology has led to a significant improvement in recognizing many languages.
Approach:	They propose to use a dataset to analyze the data available for the Hong Kong Cantonese language . they use zh-HK as a source and a state-of-the-art ASR model to build a powerful model .
Outcome:	The proposed model improves on the biggest existing dataset, Common Voice zh-HK.

Copied to clipboard

He Bai, Peng Shi, Jimmy Lin, Luchen Tan, Kun Xiong, Wen Gao, Jie Liu, Ming Li

Farewell Freebase: Migrating the SimpleQuestions Dataset to DBpedia (C18-1)

Challenge:	Experimental results show that pre-trained language model GPT2 can generate better continuations by learning to generate the in the fine-tuning stage.
Approach:	They conduct experiments on an English essay dataset using Chinese-GPT2 . they find that the model can generate better continuations by learning to generate the in the fine-tuning stage.
Outcome:	The pre-trained language model GPT2 can generate better continuations by learning to generate the in the fine-tuning stage.

Copied to clipboard

Michael Azmy, Peng Shi, Jimmy Lin, Ihab Ilyas

Video-Text Retrieval by Supervised Sparse Multi-Grained Learning (2023.findings-emnlp)

Challenge:	Existing datasets for question answering over knowledge graphs lack answer triples from Freebase . a defunct knowledge graph makes it difficult to build "real-world" question answering systems .
Approach:	They propose a benchmark dataset for simple question answering over knowledge graphs that maps SimpleQuestions entities and predicates from Freebase to DBpedia.
Outcome:	The proposed dataset provides simple yet strong baselines with and without neural networks.

Copied to clipboard

Yimu Wang, Peng Shi

Z-Code++: A Pre-trained Language Model Optimized for Abstractive Summarization (2023.acl-long)

Challenge:	Recent advances in video-text retrieval have led to improved representation learning methods.
Approach:	They propose a multi-grained sparse learning framework to learn an aligned sparsen space shared between video and text for video-text retrieval.
Outcome:	The proposed framework is superior to existing methods on video-text retrieval benchmarks.

Copied to clipboard

Pengcheng He, Baolin Peng, Song Wang, Yang Liu, Ruochen Xu, Hany Hassan, Yu Shi, Chenguang Zhu, Wayne Xiong, Michael Zeng, Jianfeng Gao, Xuedong Huang

TongGu: Mastering Classical Chinese Understanding with Knowledge-Grounded Large Language Models (2024.findings-emnlp)

Challenge:	Z-Code++ is a pre-trained language model optimized for abstractive text summarization.
Approach:	They propose a pre-trained language model optimized for abstractive text summarization that uses a two-phase pre-training technique to improve model's performance.
Outcome:	The proposed model outperforms the competing models on low-resource summarization tasks in zero-shot and few-shot settings.

Copied to clipboard

Jiahuan Cao, Dezhi Peng, Peirong Zhang, Yongxin Shi, Yang Liu, Kai Ding, Lianwen Jin

Bridging the Gap between Relevance Matching and Semantic Matching for Short Text Similarity Modeling (D19-1)

Challenge:	Large Language Models (LLMs) have demonstrated remarkable capability in Natural Language Processing (NLP), but struggle with Classical Chinese Understanding (CCU) Existing models, including general-purpose and preliminary LLMs, lack the ability to address CCU in data-demanding and knowledge-intensive tasks.
Approach:	They propose to use a classical Chinese corpora-based instruction-tuning dataset to unlock the full CCU potential of LLMs.
Outcome:	The proposed model unlocks the full CCU potential of LLMs by preserving its foundational knowledge while maintaining redundancy-aware tuning (RAT) and CCU-RAG.

Copied to clipboard

Jinfeng Rao, Linqing Liu, Yi Tay, Wei Yang, Peng Shi, Jimmy Lin

Perception, Understanding and Reasoning: A Multimodal Benchmark for Video Fake News Detection (2026.acl-long)

Challenge:	Existing techniques for relevance and semantic matching cannot be easily adapted to the other.
Approach:	They propose a model that incorporates a hybrid encoder module, a relevance matching module and co-attention mechanisms that capture context-aware semantic relatedness.
Outcome:	The proposed model incorporates a hybrid encoder module, a relevance matching module and co-attention mechanisms that capture context-aware semantic relatedness.

Copied to clipboard

Cui Yakun, Peng Qi, Fushuo Huo, Hang Du, Weijie Shi, Juntao Dai, Zhenghao Zhu, Sirui Han

Why Supervised Fine-Tuning Fails to Learn: A Systematic Study of Incomplete Learning in Large Language Models (2026.acl-long)

Challenge:	Existing video fake news detection benchmarks focus on the detection accuracy, while failing to provide fine-grained assessments for the entire detection process.
Approach:	They propose a process-oriented video fake news detection benchmark that evaluates MLLMs' perception, understanding, and reasoning capabilities in VFND.
Outcome:	The proposed model achieves sota performance on video fake news detection tasks.

Copied to clipboard

Chao Xue, Yao Wang, Mengqiao Liu, Di Liang, Xingsheng Han, Peiyang Liu, Xianjie Wu, Chenyao Lu, Lei Jiang, Yu Lu, Haibo Shi, Shuang Liang, Minlong Peng, Flora D. Salim

Cross-Lingual Training of Neural Models for Document Ranking (2020.findings-emnlp)

Challenge:	Incomplete learning is widespread and heterogeneous in large language models . authors identify five recurrent sources of incomplete learning: missing prerequisite knowledge, conflicts between SFT supervision and pre-training knowledge, internal inconsistencies within SFT data, left-side forgetting during sequential fine-tuning, and insufficient optimization for rare or complex patterns.
Approach:	They propose a diagnostic-first framework that maps incomplete learning to causes . they identify five recurrent sources of incomplete learning: missing prerequisite knowledge, conflicts between supervision and pre-training knowledge, internal inconsistencies, left-side forgetting during sequential fine-tuning, and insufficient optimization for rare or complex patterns.
Outcome:	The proposed framework maps incomplete learning to causes using observable training and inference signals.

Copied to clipboard

Peng Shi, He Bai, Jimmy Lin

Multilinguality

Challenge:	a recent study shows that multi-lingual BERT models can be used for document ranking in non-English languages . a blog post by Google suggests that the company is exploring this approach to improve web search across a number of languages.
Approach:	They propose to leverage relevance judgments in English to train neural document ranking models for mono-lingual retrieval in multiple target languages.
Outcome:	The proposed approach improves search quality in non-English languages while requiring low resources.

CoT-RAG: Integrating Chain of Thought and Retrieval-Augmented Generation to Enhance Reasoning in Large Language Models (2025.findings-emnlp)

Copied to clipboard

Feiyang Li, Peng Fang, Zhan Shi, Arijit Khan, Fang Wang, Weihao Wang, null Zhangxin-hw, Cui Yongjian

Cross-lingual Text-to-SQL Semantic Parsing with Representation Mixup (2022.findings-emnlp)

Challenge:	Chain-of-thought reasoning has two key limitations: lack of reliability when solely relying on LLM-generated reasoning chains and interference from natural language reasoning steps with the models’ inference logic.
Approach:	They propose a chain-of-thought reasoning framework with three key designs to address these issues.
Outcome:	The proposed framework improves the performance of large language models on complex tasks by incorporating knowledge graphs and learnable knowledge case-aware RAG.

Copied to clipboard

Peng Shi, Linfeng Song, Lifeng Jin, Haitao Mi, He Bai, Jimmy Lin, Dong Yu

Multilinguality

Challenge:	Experimental results show that Rex can benefit from cross-lingual training and improve the effectiveness of semantic parsers.
Approach:	They propose a Representation Mixup Framework for effectively exploiting translations in the cross-lingual Text-to-SQL task.
Outcome:	The proposed framework can benefit from cross-lingual training and improve the effectiveness of semantic parsers, achieving state-of-the-art performance.

You Only Read Once (YORO): Learning to Internalize Database Knowledge for Text-to-SQL (2025.naacl-long)

Copied to clipboard

Hideo Kobayashi, Wuwei Lan, Peng Shi, Shuaichen Chang, Jiang Guo, Henghui Zhu, Zhiguo Wang, Patrick Ng

Programming Languages in NLP

Challenge:	Existing text-to-SQL systems encode the same schema for every question, resulting in unnecessary high inference cost and missing crucial database knowledge.
Approach:	They propose a paradigm that directly internalizes database knowledge into the parametric knowledge of a text-to-SQL model during training and eliminates the need for schema encoding during inference.
Outcome:	The proposed paradigm significantly reduces the input token length by 66%-98% and outperforms traditional systems on three benchmarks.

PACE: Prefix-Protected and Difficulty-Aware Compression for Efficient Reasoning (2026.findings-acl)

Copied to clipboard

Ruixiang Feng, Yuntao Wen, Silin Zhou, Ke Shi, Yifan Wang, Ran Le, Zhenwei An, Zongchao Chen, Chen Yang, Guangyue Peng, Yiming Jia, Dongsheng Wang, Tao Zhang, Lisi Chen, Yang Song, Shen Gao, Shuo Shang

RECALL: REpresentation-aligned Catastrophic-forgetting ALLeviation via Hierarchical Model Merging (2025.emnlp-main)

Challenge:	Existing LRMs often suffer from "overthinking" and excessively long reasoning traces . a dual-level framework for length compression of LRM is proposed .
Approach:	They propose a framework for prefix-protected and difficulty-aware compression under hierarchical supervision.
Outcome:	The proposed framework reduces token usage while improving accuracy on math benchmarks.

Copied to clipboard

Bowen Wang, Haiyuan Wan, Liwen Shi, Chen Yang, Peng He, Yue Ma, Haochen Han, Wenhao Li, Tiao Tan, Yongjian Li, Fangming Liu, Gong Yifan, Sheng Zhang

ToolPRM: Fine-Grained Inference Scaling of Structured Outputs for Function Calling (2026.acl-long)

Challenge:	Existing models that require task labels or performance trade-offs are susceptible to catastrophic forgetting.
Approach:	They propose a representation-aware model merging framework for continual learning without access to historical data.
Outcome:	The proposed framework outperforms baselines in knowledge retention and generalization across five NLP tasks and multiple continual learning scenarios.

Copied to clipboard

Jianghao Lin, Yuanyuan Shi, Xin Peng, Renjie Ding, Hairui Wang, Yuxuan Peng, Bizhe Bai, Weixi Song, Fengshuo Bai, Huacan Chai, Weinan Zhang, Fei Huang, Ying Wen

Challenge:	Existing research on inference scaling focuses on unstructured output generation tasks, such as mathematical problems.
Approach:	They propose an inference-scaling framework that combines fine-grained beam search with ToolPRM, a process reward model scoring each intra-call decision.
Outcome:	The proposed framework outperforms outcome and coarse-grained reward models in predictive accuracy and yields consistent test-time gains on multiple function-calling benchmarks.

ESPnet-ST-v2: Multipurpose Spoken Language Translation Toolkit (2023.acl-demo)

Copied to clipboard

Brian Yan, Jiatong Shi, Yun Tang, Hirofumi Inaguma, Yifan Peng, Siddharth Dalmia, Peter Polák, Patrick Fernandes, Dan Berrebbi, Tomoki Hayashi, Xiaohui Zhang, Zhaoheng Ni, Moto Hira, Soumi Maiti, Juan Pino, Shinji Watanabe

Text Generation

Challenge:	ESPnet-ST-v2 is a revamp of the open-source spoken language translation toolkit . it supports offline speech-to-text translation (ST), simultaneous speech- to-text (SST), and offline speech to-speech (S2ST)
Approach:	They propose to revamp the open-source ESPnet-ST toolkit to support offline speech-to-text translation, simultaneous speech- to-text and offline speech to-speech translation.
Outcome:	The updated version of ESPnet-ST supports offline speech-to-text translation (ST), simultaneous speech- to-text (SST), and offline speech to-speech translation (S2ST).

MulDimIF: A Multi-Dimensional Constraint Framework for Evaluating and Improving Instruction Following in Large Language Models (2026.findings-acl)

Copied to clipboard

Junjie Ye, Caishuang Huang, Zhuohan Chen, Wenjie Fu, Chenyuan Yang, Leyi Yang, Yilong Wu, Peng Wang, Meng Zhou, Xiaolong Yang, Tao Gui, Qi Zhang, Zhongchao Shi, Jianping Fan, Xuanjing Huang

Praetor: A Fine-Grained Generative LLM Evaluator with Instance-Level Customizable Evaluation Criteria (2025.acl-long)

Challenge:	Existing research has focused on constraint categories, offering little guidance for improving instruction following abilities.
Approach:	They propose a multi-dimensional constraint framework that allows for instruction following . they construct 9,106 code-verifiable samples and evaluate 18 LLMs .
Outcome:	The proposed framework improves instruction following performance without compromising general performance.

Copied to clipboard

Yongqi Leng, Renren Jin, Yue Chen, Zhuowen Han, Ling Shi, Jianxiang Peng, Lei Yang, Juesi Xiao, Deyi Xiong

Challenge:	Existing evaluation methods are inadequate to evaluate large language models (LLMs).
Approach:	They propose a fine-grained generative LLM evaluator with instance-level customazable evaluation criteria that can be used to evaluate large language models.
Outcome:	The proposed model outperforms existing LLM evaluators and instruction-tuned LLMs on multiple benchmarks and sets new SOTA results.

Reason Only When Needed: Efficient Generative Reward Modeling via Model-Internal Uncertainty (2026.findings-acl)

Copied to clipboard

Chao Xue, Yao Wang, Mengqiao Liu, Di Liang, Xingsheng Han, Peiyang Liu, Xianjie Wu, Chenyao Lu, Lei Jiang, Yu Lu, Haibo Shi, Shuang Liang, Minlong Peng, Flora D. Salim

From Curated Data to Scalable Models: Continual Pre-training of Dense and MoE Large Language Models for Tibetan (2026.acl-long)

Challenge:	Existing approaches to generating reward models rely on voting-based mechanisms to evaluate CoT outputs.
Approach:	They propose an efficient generative reward modeling framework grounded in model-internal uncertainty.
Outcome:	The proposed framework reduces inference cost while improving answer accuracy.

Copied to clipboard

Lei Yang, Leiyu Pan, Bojian Xiong, Renren Jin, Shaowei Zhang, Yue Chen, Ling Shi, Jiang Zhou, Junru Wu, Zhen Wang, Jianxiang Peng, Juesi Xiao, Tianyu Dong, Zhuowen Han, Zhuo Chen, Yuqi Ren, Deyi Xiong

RAP-ID: Mechanistic Prompt Injection Detection via Impostor Behavior Analysis (2026.findings-acl)

Challenge:	Large language models have achieved remarkable success across a wide range of tasks, yet their performance remains heavily biased toward high-resource languages.
Approach:	They propose a pipeline for advancing Tibetan language modeling through multilingual continual pre-training with Tibetan, Chinese, and English.
Outcome:	The proposed model outperforms open-source and Tibetan-focused models on diverse tasks.

Copied to clipboard

Yuchen Yang, Lei Peng, Yujie He, Yang yu, Zhongxin Wu, Yanlei Shi

Nature-Inspired Population-Based Evolution of Large Language Models (2026.acl-long)

Challenge:	Existing defenses rely on externally deployed guardrail models or response inspection . current defenses depend on external guardrails or response inspecting .
Approach:	They propose a mechanistic, train-free detection framework that operates exclusively on internal state dynamics during the initial forward pass.
Outcome:	The proposed framework achieves competitive performance with significant overall improvements compared to heuristic methods.

Copied to clipboard

Yiqun Zhang, Peng Ye, Xiaocui Yang, Shi Feng, Shufei Zhang, Lei Bai, Wanli Ouyang, Shuyue Hu

SWE-QA: Can Language Models Answer Repository-level Code Questions? (2026.findings-acl)

Challenge:	a new framework for population-based evolution of large language models is emerging . a population-driven evolution of LLMs is a key component of evolution, authors say .
Approach:	They propose a framework that allows for population-based evolution of large language models . they start with a population of parent LLMs and allow this population to evolve .
Outcome:	The proposed framework outperforms existing methods on 12 datasets.

Copied to clipboard

Weihan Peng, Yuling Shi, Yuhang Wang, Xinyun Zhang, Beijun Shen, Xiaodong Gu

DiplomacyAgent: Do LLMs Balance Interests and Ethical Principles in International Events? (2025.emnlp-main)

Challenge:	Existing benchmarks for understanding and reasoning about entire soft-ware repositories focus on small, self-contained code snippets.
Approach:	They propose a repository-level code question answering benchmark to facilitate research on automated QA systems in real-world repositories.
Outcome:	The proposed benchmarks are designed to facilitate research on automated QA systems in real-world repositories.

Copied to clipboard

Jianxiang Peng, Ling Shi, Xinwei Wu, Hanwen Zhang, Fujiang Liu, Haocheng Lyu, Deyi Xiong

Logic-Consistency Text Generation from Semantic Parses (2021.findings-acl)

Challenge:	a new study examines the safety implications of large language models in diplomatic positions . it identifies potential risks and ideological biases that could arise from LLMs .
Approach:	They propose an LLM-based multi-agent system for diplomatic position analysis . they propose ethical constraint measures to enhance the safety of LLMs .
Outcome:	The proposed system assesses the safety implications of large language models in diplomacy . it reveals that LLMs could exhibit a strong bias towards interests, leading to unsafe decisions .

Copied to clipboard

Chang Shu, Yusen Zhang, Xiangyu Dong, Peng Shi, Tao Yu, Rui Zhang

Text Generation

Challenge:	Text generation from semantic parses is challenging due to the complexity of the inner logic and the lack of automatic evaluation metrics for logic consistency.
Approach:	They propose a framework for logic consistent text generation from semantic parses that employs iterative training procedures and quality control.
Outcome:	The proposed framework enhances logic consistency and human evaluation on two benchmark datasets.

Better Language Model with Hypernym Class Prediction (2022.acl-long)

Copied to clipboard

He Bai, Tong Wang, Alessandro Sordoni, Peng Shi

Strong Baselines for Simple Question Answering over Knowledge Graphs with and without Neural Networks (N18-2)

Challenge:	Class-based language models (LMs) have been devised to address context sparsity in n-gram LMs for decades.
Approach:	They propose to use class-based prediction to improve generalization for rare words by annealing from predicting the class to token prediction during training.
Outcome:	The proposed model improves perplexity without sacrificing performance on rare words.

Copied to clipboard

Salman Mohammed, Peng Shi, Jimmy Lin

ReCUT: Balancing Reasoning Length and Accuracy in LLMs via Stepwise Trails and Preference Optimization (2025.findings-emnlp)

Challenge:	Existing work on simple question answering over knowledge graphs involves increasingly complex NN architectures.
Approach:	They propose to decompose the problem into entity detection, entity linking, relation prediction, evidence combination and heuristics.
Outcome:	The proposed approach outperforms existing models and benchmarks on a simple QA task.

Copied to clipboard

Zhensheng Jin, Xinze Li, Yifan Ji, Chunyi Peng, Zhenghao Liu, Qi Shi, Yukun Yan, Shuo Wang, Furong Peng, Ge Yu

Reasoning

Challenge:	Existing methods to train LLMs suffer from overthinking, leading to lengthy reasoning traces . Existing approaches to train large language models suffer from this problem .
Approach:	They propose a method to combine multiple reasoning chains for training LLMs . they use stepwise exploration and long-short switched sampling to evaluate reasoning paths .
Outcome:	The proposed method reduces reasoning lengths by approximately 30-50% . it also maintains or improves reasoning accuracy compared to baselines .

Improving Chinese Pop Song and Hokkien Gezi Opera Singing Voice Synthesis by Enhancing Local Modeling (2023.emnlp-main)

Copied to clipboard

Peng Bai, Yue Zhou, Meizhen Zheng, Wujin Sun, Xiaodong Shi

Speech & Audio in NLP

Challenge:	Singing Voice Synthesis (SVS) synthesizes pleasing vocals based on music scores and lyrics . current acoustic models ignore the significance of local modeling within the sequence and the hard-to-synthesize parts in the predicted mel-spectrogram .
Approach:	They propose a method to enhance local modeling in the acoustic model by focusing on phoneme tokens located before and after the phoneme.
Outcome:	The proposed method improves local modeling in the acoustic model by focusing on the hard-to-synthesize parts of the predicted mel-spectrogram.

Unified Low-Resource Sequence Labeling by Sample-Aware Dynamic Sparse Finetuning (2023.emnlp-main)

Copied to clipboard

Sarkar Snigdha Sarathi Das, Haoran Zhang, Peng Shi, Wenpeng Yin, Rui Zhang

CI-AVSR: A Cantonese Audio-Visual Speech Datasetfor In-car Command Recognition (2022.lrec-1)

Challenge:	Named Entity Recognition, Relation Extraction, Semantic Role Labeling are examples of sequence labeling problems that require finetuning to the target format.
Approach:	They propose a dynamic sparse finetuning strategy that selectively focuses on a fraction of parameters, informed by feedback from highly regressing examples.
Outcome:	The proposed approach improves performance in low-resource settings and in extreme low-level settings.

Copied to clipboard

Wenliang Dai, Samuel Cahyawijaya, Tiezheng Yu, Elham J. Barezi, Peng Xu, Cheuk Tung Yiu, Rita Frieske, Holy Lovenia, Genta Winata, Qifeng Chen, Xiaojuan Ma, Bertram Shi, Pascale Fung

Prompting Large Language Models to Tackle the Full Software Development Lifecycle: A Case Study (2025.coling-main)

Challenge:	In-car smart assistants should be able to process general as well as car-related commands and perform corresponding actions, which eases driving and improves safety.
Approach:	They propose a dataset for in-car command recognition in the cantonese language with both video and audio data.
Outcome:	The proposed model can achieve a considerable quality on the clean test set, but the speech recognition quality on noisy data is still inferior.

Copied to clipboard

Bowen Li, Wenhan Wu, Ziwei Tang, Lin Shi, John Yang, Jinyang Li, Shunyu Yao, Chen Qian, Binyuan Hui, Qicheng Zhang, Zhiyin Yu, He Du, Ping Yang, Dahua Lin, Chao Peng, Kai Chen

ASCEND: A Spontaneous Chinese-English Dataset for Code-switching in Multi-turn Conversation (2022.lrec-1)

Challenge:	Existing benchmarks focused on simplified or isolated aspects of coding, ignoring the full spectrum of programming challenges.
Approach:	They propose a case study that examines the performance of large language models across the entire software development lifecycle with four programming languages, multiple domains, and carefully designed and verified metrics for each task.
Outcome:	The proposed model performs across the entire software development lifecycle, including design, environment setup, implementation, acceptance testing, and unit testing.

Copied to clipboard

Holy Lovenia, Samuel Cahyawijaya, Genta Winata, Peng Xu, Yan Xu, Zihan Liu, Rita Frieske, Tiezheng Yu, Wenliang Dai, Elham J. Barezi, Qifeng Chen, Xiaojuan Ma, Bertram Shi, Pascale Fung

Dialogue Systems & Conversational Agents

Challenge:	Code-switching is a speech phenomenon occurring when a speaker switches language during a conversation.
Approach:	They propose to collect Mandarin Chinese-English code-switching corpus from read speech rather than spontaneous speech to address this phenomenon.
Outcome:	ASCEND consists of 10.62 hours of clean speech, collected from 23 bilingual speakers of Chinese and English.

Simple Attention-Based Representation Learning for Ranking Short Social Media Posts (N19-1)

Copied to clipboard

Peng Shi, Jinfeng Rao, Jimmy Lin

Representation Learning

Challenge:	Existing approaches to ranking short social media posts are complex and require different components to capture a multitude of relevance signals.
Approach:	They propose a word-level Siamese architecture with attention-based mechanisms for capturing semantic "soft" matches between query and post tokens.
Outcome:	The proposed model is faster and simpler than existing models and more efficient than existing approaches.

Construction of Paired Knowledge Graph - Text Datasets Informed by Cyclic Evaluation (2024.lrec-main)

Copied to clipboard

Ali Mousavi, Xin Zhan, He Bai, Peng Shi, Theodoros Rekatsinas, Benjamin Han, Yunyao Li, Jeffrey Pound, Joshua M. Susskind, Natalie Schluter, Ihab F. Ilyas, Navdeep Jaitly

Structured Data in NLP

Challenge:	Prior studies have shown that sequence-to-sequence models learn to hallucinate when the conditioning data has poor correlation with the sequence being produced.
Approach:	They construct a dataset that pairs Knowledge Graphs (KG) and text together and compare their results to a cyclic evaluation model.
Outcome:	The proposed model performs better on cyclic generation of KGs than on KG-T, but less well on synchronization of KTs.

Towards Robust Speech Representation Learning for Thousands of Languages (2024.emnlp-main)

Copied to clipboard

William Chen, Wangyou Zhang, Yifan Peng, Xinjian Li, Jinchuan Tian, Jiatong Shi, Xuankai Chang, Soumi Maiti, Karen Livescu, Shinji Watanabe

Robustness in NLP

Challenge:	XEUS is a cross-lingual encoder for universal speech that can be trained on 1 million hours of data across 4057 languages.
Approach:	They propose a Cross-lingual Encoder for Universal Speech that can be trained on 1 million hours of data across 4057 languages and a newly created corpus of 7400+ hours from 4057 .
Outcome:	The proposed model outperforms state-of-the-art models on several benchmarks and outperfies MMS 1B and w2v-BERT 2.0 v2 by 0.8% and 4.4% respectively.

TL-Training: A Task-Feature-Based Framework for Training Large Language Models in Tool Use (2025.findings-emnlp)

Copied to clipboard

Junjie Ye, Yilong Wu, Sixian Li, Yuming Yang, Zhiheng Xi, Tao Gui, Qi Zhang, Xuanjing Huang, Peng Wang, Zhongchao Shi, Jianping Fan, Zhengyin Du

Robustness in NLP

Challenge:	a new approach to training large language models (LLMs) overlooks task-specific characteristics in tool use, leading to performance bottlenecks.
Approach:	They propose a task-feature-based framework that mitigates the effects of suboptimal training data . they use a dataset to train large-scale LLMs and a reward mechanism tailored to error categories .
Outcome:	The proposed framework matches or surpasses open- and closed-source LLMs in tool-use performance using only 1,217 training data points.

Web Fraud Attacks Against LLM-Driven Multi-Agent Systems (2026.findings-acl)

Copied to clipboard

Dezhang Kong, Hujin Peng, Yilun Zhang, Lele Zhao, Zhenhua Xu, Shi Lin, Changting Lin, Meng Han

Robustness in NLP

Challenge:	Large Language Model (LLM)-driven multi-agent systems (MAS) are rapidly gaining popularity, and its inherent security risks are rapidly becoming a concern.
Approach:	They propose a novel attack manipulating unique structures of web links to deceive MAS by using homoglyph deception, sub-directory nesting, and parameter obfuscation.
Outcome:	The proposed attacks exploit unique structures of web links to deceive MAS . they exhibit significant destructive potential across different MAS architectures .

XRICL: Cross-lingual Retrieval-Augmented In-Context Learning for Cross-lingual Text-to-SQL Semantic Parsing (2022.findings-emnlp)

Copied to clipboard

Peng Shi, Rui Zhang, He Bai, Jimmy Lin

Dense Procedure Captioning in Narrated Instructional Videos (P19-1)

Challenge:	Existing work focuses on English datasets, and it is unclear whether large language models can serve as competitive semantic parsers for other languages.
Approach:	They propose a framework that learns to retrieve relevant English exemplars for a given query to construct prompts.
Outcome:	The proposed framework learns to retrieve relevant English exemplars for a given query to construct prompts.

Copied to clipboard

Botian Shi, Lei Ji, Yaobo Liang, Nan Duan, Peng Chen, Zhendong Niu, Ming Zhou

UnifiedSKG: Unifying and Multi-Tasking Structured Knowledge Grounding with Text-to-Text Language Models (2022.emnlp-main)

Challenge:	Existing models for video dense captioning learn video segments and generate captions without considering transcripts.
Approach:	They propose a model to generate procedure captions from narrated instructional videos . they extract procedures by a cross-modality module and generate captions by encoding video frames and transcripts within each extracted procedure.
Outcome:	The proposed model can extract procedures from narrated instructional videos and generate procedure captions by encoding video frames and transcripts.

Copied to clipboard

Tianbao Xie, Chen Henry Wu, Peng Shi, Ruiqi Zhong, Torsten Scholak, Michihiro Yasunaga, Chien-Sheng Wu, Ming Zhong, Pengcheng Yin, Sida I. Wang, Victor Zhong, Bailin Wang, Chengzu Li, Connor Boyle, Ansong Ni, Ziyu Yao, Dragomir Radev, Caiming Xiong, Lingpeng Kong, Rui Zhang, Noah A. Smith, Luke Zettlemoyer, Tao Yu

SPEAK: Spiking Neurons as an Entropy-Aware Tokenizer for Large Language Models (2026.acl-long)

Challenge:	Structured knowledge grounding (SKG) uses structured knowledge to complete user requests . since inputs and outputs of SKG tasks are heterogeneous, they have been studied separately .
Approach:	They propose a framework that unifies 21 SKG tasks into a text-to-text format . they use unifiedSKG to benchmark T5 with different sizes .
Outcome:	The proposed framework unifies 21 SKG tasks into a text-to-text format . it achieves state-of-the-art performance on almost all of the 21 tasks, the authors show .

Copied to clipboard

Ming Chen, Wenyao Li, Chao Liang, Shi Gu, Peng Lin, De Ma, Huajin Tang, Qian Zheng, Gang Pan

Calibrating the Confidence of Large Language Models by Eliciting Fidelity (2024.emnlp-main)

Challenge:	Existing tokenizers fail to explicitly leverage historical tokenization results . large language models (LLMs) have demonstrated remarkable effectiveness across NLP tasks .
Approach:	They propose a tokenizer that integrates spiking neurons to explicitly leverage historical tokenization results.
Outcome:	The proposed tokenizer leverages historical tokenization results, but does not selectively leverage history based on contextual relevance.

Copied to clipboard

Mozhi Zhang, Mianqiu Huang, Rundong Shi, Linsen Guo, Chong Peng, Peng Yan, Yaqian Zhou, Xipeng Qiu

Aligning Cross-Lingual Entities with Multi-Aspect Information (D19-1)

Challenge:	Large language models with RLHF and RLAIF have good alignment but exhibit overconfidence post-alignment.
Approach:	They propose a plug-and-play method to estimate the confidence of large language models.
Outcome:	The proposed method has shown good calibration performance on 6 RLHF-LMs on four MCQA datasets.

Copied to clipboard

Hsiu-Wei Yang, Yanyan Zou, Peng Shi, Wei Lu, Jimmy Lin, Xu Sun

Analyzing the Effects of Supervised Fine-Tuning on Model Knowledge from Token and Parameter Levels (2025.emnlp-main)

Challenge:	Existing knowledge graphs that represent entities in different languages are not covered by existing systems.
Approach:	They propose two ways to embed entities from multilingual knowledge graphs into the same vector space, where equivalent entities are close to each other.
Outcome:	The proposed method significantly outperforms existing systems on two benchmark datasets.

Copied to clipboard

Junjie Ye, Yuming Yang, Yang Nan, Shuo Li, Qi Zhang, Tao Gui, Xuanjing Huang, Peng Wang, Zhongchao Shi, Jianping Fan