Papers by Haoran Li

PrivacyRestore: Privacy-Preserving Inference in Large Language Models via Privacy Removal and Restoration (2025.acl-long)

Challenge:	Existing models suffer from temporal redundancy when leveraged under dynamic settings.
Approach:	They propose a temporal knowledge graph extrapolation method which solves temporal redundancy issues by using cyclic rules to capture more information lurking in TKGs.
Outcome:	The proposed model captures more information lurking in TKGs, and also mines and properly leverages acyclic rules, which has not been explored by existing models.

Copied to clipboard

Ziqian Zeng, Jianwei Wang, Junyao Yang, Zhengdong Lu, Haoran Li, Huiping Zhuang, Cen Chen

Recurrent Attention Networks for Long-text Modeling (2023.findings-acl)

Challenge:	Existing privacy protection methods for large language models suffer from performance degradation or large inference time overhead.
Approach:	They propose a plug-and-play method to protect the privacy of user inputs during LLM inference . they use offline restoration vectors to train restoration vector for each privacy span type .
Outcome:	The proposed method can prevent the linear growth of the privacy budget.

Copied to clipboard

Xianming Li, Zongxi Li, Xiaotian Luo, Haoran Xie, Xing Lee, Yingbin Zhao, Fu Lee Wang, Qing Li

Information Extraction & Text Mining

Challenge:	Existing approaches to encoding long documents using self-attention have been limited by quadratic computational complexities and limited application in long text processing.
Approach:	They propose a long-document encoding model that allows the recurrent operation of self-attention.
Outcome:	The proposed model extracts global semantics in token-level and document-level representations, making it inherently compatible with both sequential and sequential tasks.

Condensing Multilingual Knowledge with Lightweight Language-Specific Modules (2023.emnlp-main)

Copied to clipboard

Haoran Xu, Weiting Tan, Shuyue Li, Yunmo Chen, Benjamin Van Durme, Philipp Koehn, Kenton Murray

Green & Sustainable NLP

Challenge:	Existing methods to boost performance in multilingual models but scalability is difficult to manage.
Approach:	They propose a method that incorporates language-specific (LS) modules to boost model performance.
Outcome:	The proposed method outperforms state-of-the-art methods while outperforming existing methods.

Sentence Embedding Leaks More Information than You Expect: Generative Embedding Inversion Attack to Recover the Whole Sentence (2023.findings-acl)

Copied to clipboard

Haoran Li, Mingshi Xu, Yangqiu Song

Defense Against Prompt Injection Attack by Leveraging Attack Techniques (2025.acl-long)

Challenge:	Sentence-level representations are beneficial for various natural language processing tasks.
Approach:	They propose a generative embedding inversion attack that reconstructs input sequences based only on their sentence embeddeds.
Outcome:	The proposed model outperforms previous embedding inversion attacks in classification metrics and generates coherent and contextually similar sentences as original inputs.

Copied to clipboard

Yulin Chen, Haoran Li, Zihao Zheng, Dekai Wu, Yangqiu Song, Bryan Hooi

Advancing Collaborative Debates with Role Differentiation through Multi-Agent Reinforcement Learning (2025.acl-long)

Challenge:	Recent attacks leverage LLMs’ instruction-following abilities and their inabilities to distinguish instructions injected in the data content.
Approach:	They invert the intention of prompt injection methods to develop novel defense methods based on previous training-free attack methods by repeating the attack process with the original input instruction rather than the injected instruction.
Outcome:	The proposed methods outperform existing defense approaches, achieving state-of-the-art results.

Copied to clipboard

Haoran Li, Ziyi Su, Yun Xue, Zhiliang Tian, Yiping Song, Minlie Huang

Representation Learning

Challenge:	Multi-agent collaborative tasks exhibit exceptional capabilities in natural language applications and generation.
Approach:	They propose a multi-LLM Cooperation framework with automatic role assignment capabilities that allows multiple agents to embed roles in turn-based speaking.
Outcome:	The proposed framework improves collaboration and expertise among agents and teams by enabling them to share roles and develop complementary strengths from the optimization level.

TopicAttack: An Indirect Prompt Injection Attack via Topic Transition (2025.emnlp-main)

Copied to clipboard

Yulin Chen, Haoran Li, Yuexin Li, Yue Liu, Yangqiu Song, Bryan Hooi

You Don’t Know My Favorite Color: Preventing Dialogue Representations from Revealing Speakers’ Private Personas (2022.naacl-main)

Challenge:	Recent attacks have demonstrated potential, but their abrupt instruction injection often undermines their effectiveness.
Approach:	They propose a method that prompts the LLM to generate a fabricated conversational transition prompt that gradually shifts the topic toward the injected instruction.
Outcome:	The proposed method achieves state-of-the-art performance with an attack success rate (ASR) over 90% in most cases, even when various defense methods are applied.

Copied to clipboard

Haoran Li, Yangqiu Song, Lixin Fan

On the Role of Entity and Event Level Conceptualization in Generalizable Reasoning: A Survey of Tasks, Methods, Applications, and Future Directions (2025.findings-emnlp)

Challenge:	Social chatbots evolve rapidly with large pretrained language models.
Approach:	They propose effective defense objectives to protect persona leakage from hidden states by a simple neural network.
Outcome:	The proposed defense objectives reduce the attack accuracy from 37.6% to 0.5% while preserving language models’ powerful generation ability.

Copied to clipboard

Weiqi Wang, Tianqing Fang, Haochen Shi, Baixuan Xu, Wenxuan Ding, Liyu Zhang, Wei Fan, Jiaxin Bai, Haoran Li, Xin Liu, Yangqiu Song

Reasoning

Challenge:	Conceptualization is a fundamental element of human cognition and plays a pivotal role in generalizable reasoning.
Approach:	They propose to categorize different types of conceptualizations into four levels based on the types of instances being conceptualized.
Outcome:	The proposed categorization of different types of conceptualizations into four levels based on the types of instances being conceptualized .

CondAmbigQA: A Benchmark and Dataset for Conditional Ambiguous Question Answering (2025.emnlp-main)

Copied to clipboard

Zongxi Li, Yang Li, Haoran Xie, S. Joe Qin

Reasoning

Challenge:	Large language models (LLMs) generate unreliable responses due to their cognitive alignment of context and intent.
Approach:	They propose a benchmark to identify possible implicit assumptions in QA questions . they use retrieved Wikipedia fragments to identify interpretations for a given query .
Outcome:	The proposed benchmark identifies possible implicit assumptions and improves answer accuracy by 11.75% . retrieved Wikipedia fragments help identify possible interpretations for a given query .

Context Reasoner: Incentivizing Reasoning Capability for Contextualized Privacy and Safety Compliance via Reinforcement Learning (2025.emnlp-main)

Copied to clipboard

Wenbin Hu, Haoran Li, Huihao Jing, Qi Hu, Ziqian Zeng, Sirui Han, Xu Heli, Tianshu Chu, Peizhao Hu, Yangqiu Song

Robustness via Referencing: Defending against Prompt Injection Attacks by Referencing the Executed Instruction (2026.findings-acl)

Challenge:	Current mitigation strategies fail to preserve contextual reasoning capabilities in risky scenarios, leading to systemic risks for legal compliance.
Approach:	They propose to use reinforcement learning with a rule-based reward to incentivize contextual reasoning capabilities while enhancing compliance with safety and privacy norms.
Outcome:	The proposed model outperforms Qwen2.5-7B-Instruct model in safety and privacy benchmarks and achieves +8.58% accuracy improvement.

Copied to clipboard

Yulin Chen, Haoran Li, Yuan Sui, Yue Liu, Yufei He, Xiaoling Bai, Chi Fei, Li Yabo, Haozhe Ma, Yangqiu Song, Bryan Hooi

iMOVE : Instance-Motion-Aware Video Understanding (2025.findings-acl)

Challenge:	Prompt injection attacks manipulate large language models (LLMs) by misleading them to deviate from the original input instructions and execute maliciously injected instructions.
Approach:	They propose a prompt injection defense method that suppresses the model's instruction-following tendencies rather than suppressing them.
Outcome:	The proposed method outperforms prompt-engineering-based approaches and fine-tuning methods and reduces the ASR to nearly 0% in some scenarios.

Copied to clipboard

Jiaze Li, Yaya Shi, Zongyang Ma, Haoran Xu, Yandong.bai Yandong.bai, Huihui Xiao, Ruiwen Kang, Fan Yang, Tingting Gao, Di Zhang

Visual Data in NLP

Challenge:	Recent advances in Video Large Language Models have led to rapid development, significantly enhancing the capture of overall video semantics and achieving remarkable performance in general video understanding tasks.
Approach:	They propose a large-scale instance-motion-aware video instruction-tuning dataset iMOVE that utilizes Event-awful Spatiotemporal Efficient Modeling to retain informative instance spatiotemporal motion details while maintaining computational efficiency.
Outcome:	The proposed model excels in video temporal understanding and general video understanding.

JDDC 2.1: A Multimodal Chinese Dialogue Dataset with Joint Tasks of Query Rewriting, Response Generation, Discourse Parsing, and Summarization (2022.emnlp-main)

Copied to clipboard

Nan Zhao, Haoran Li, Youzheng Wu, Xiaodong He

Federated LoRA Fine-Tuning with Pipelined Error-Mitigated Aggregation and Matrix-Wise Freezing (2026.findings-acl)

Challenge:	e-commerce users express their needs using text, images, or videos . but detailed information provided by images is limited, and customer service systems cannot understand the intent of users without the input text.
Approach:	They construct a large-scale multimodal multi-turn dialogue dataset from a mainstream Chinese E-commerce platform . the dataset contains about 246K dialogue sessions, 3M utterances, and 507K images .
Outcome:	The proposed dataset contains 246K dialogue sessions, 3M utterances, 507K images . it also includes product knowledge bases and image category annotations .

Copied to clipboard

Haoran Wang, Xiong Wang, Yuqing Li, Jing Chen, Junyi Zhang, Nan Yan, Kun He, Wei Wang

Can Indirect Prompt Injection Attacks Be Detected and Removed? (2025.acl-long)

Challenge:	Existing methods for fine-tuning large language models often suffer from biased model aggregation and are hindered by significant communication and computation burden.
Approach:	They propose a Federated low-rank adaptation system for large language models that leverages pipelined error-mitigated model aggregation and adaptive matrix-wise parameter freezing to mitigate aggregations.
Outcome:	The proposed system improves time-to-target by 2.17-8.48 on real-world datasets.

Copied to clipboard

Yulin Chen, Haoran Li, Yuan Sui, Yufei He, Yue Liu, Yangqiu Song, Bryan Hooi

Tuna: Instruction Tuning using Feedback from Large Language Models (2023.findings-emnlp)

Challenge:	Recent studies have developed various detection mechanisms to protect against prompt injection attacks.
Approach:	They investigate the feasibility of detecting and removing indirect prompt injection attacks . they use two methods to evaluate their performance and train detection models .
Outcome:	The proposed method is based on a benchmark dataset and is available on github . it evaluates the performance of existing models and open-source detection models .

Copied to clipboard

Haoran Li, Yiran Liu, Xingxing Zhang, Wei Lu, Furu Wei

Syntax-augmented Multilingual BERT for Cross-lingual Transfer (2021.acl-long)

Challenge:	LLms like LLaMA have shown to be cost-effective for generating better responses . however, the instruction-tuned model has only seen one response per instruction .
Approach:	They propose to fine tune an instruction-tuned LLM using probabilistic ranking and contextual ranking approaches to increase the likelihood of generating better responses.
Outcome:	The proposed model improves on Super Natural Instructions, LMentry and Vicuna QA.

Copied to clipboard

Wasi Ahmad, Haoran Li, Kai-Wei Chang, Yashar Mehdad

Multilinguality

Challenge:	Existing studies show that pre-trained multilingual text encoders capture language syntax, helping cross-lingual transfer.
Approach:	They provide language syntax and train mBERT to encode universal dependency tree structure.
Outcome:	The proposed model improves cross-lingual transfer on PAWS-X and MLQA benchmarks by 1.4 and 1.6 points on average across all languages.

Evaluating Readability and Faithfulness of Concept-based Explanations (2024.emnlp-main)

Copied to clipboard

Meng Li, Haoran Jin, Ruixuan Huang, Zhihao Xu, Defu Lian, Zijia Lin, Di Zhang, Xiting Wang

Explainability & Interpretability in NLP

Challenge:	Existing methods for evaluating concepts from different perspectives lack a unified formalization.
Approach:	They propose a formal definition of concepts generalizing to diverse concept-based explanations’ settings and apply it to other types of explanations or tasks.
Outcome:	Extensive experimental analysis was carried out to determine the evaluation measures for explanation evaluation measures.

Sub-network Discovery and Soft-masking for Continual Learning of Mixed Tasks (2023.findings-emnlp)

Copied to clipboard

Zixuan Ke, Bing Liu, Wenhan Xiong, Asli Celikyilmaz, Haoran Li

K-PLUG: Knowledge-injected Pre-trained Language Model for Natural Language Understanding and Generation in E-Commerce (2021.findings-emnlp)

Challenge:	Existing methods for Continual Learning (CL) have limited KT and catastrophic forgetting . a new method overcomes CF by isolating the knowledge of each task .
Approach:	They propose a method to overcome catastrophic forgetting and encourage knowledge transfer . they propose to discover a sub-network for each task and a soft-masking mechanism to preserve the previous knowledge.
Outcome:	The proposed method outperforms baselines in classification, generation, information extraction and their mixture.

Copied to clipboard

Song Xu, Haoran Li, Peng Yuan, Yujia Wang, Youzheng Wu, Xiaodong He, Ying Liu, Bowen Zhou

AlphaOne: Reasoning Models Thinking Slow and Fast at Test Time (2025.emnlp-main)

Challenge:	Existing pre-trained language models are not explicitly aware of domain-specific knowledge, which is essential for downstream tasks in many domains, such as tasks in e-commerce scenarios.
Approach:	They propose a knowledge-injected pre-trained language model that can be transferred to both natural language understanding and generation tasks.
Outcome:	The proposed model significantly outperforms baselines across the board in e-commerce scenarios.

Copied to clipboard

Junyu Zhang, Runpei Dong, Han Wang, Xuying Ning, Haoran Geng, Peihao Li, Xialin He, Yutong Bai, Jitendra Malik, Saurabh Gupta, Huan Zhang

Contrastive Representation Learning for Exemplar-Guided Paraphrase Generation (2021.findings-emnlp)

Challenge:	Existing monotonic scaling methods for large reasoning models are not reliable.
Approach:	They propose a universal framework for modulating reasoning progress in large reasoning models at test time.
Outcome:	The proposed framework unifies and generalizes existing monotonic scaling methods and enables flexible and dense slow-to-fast reasoning modulation.

Copied to clipboard

Haoran Yang, Wai Lam, Piji Li

Representation Learning

Challenge:	Exemplar-Guided Paraphrase Generation (EGPG) aims to generate a target sentence which conforms to the style of the given exemplar while encapsulating the content information of the source sentence.
Approach:	They propose a method to generate a target sentence which conforms to the style of the given exemplar while encapsulating the content information of the source sentence.
Outcome:	The proposed method can generate a target sentence which conforms to the style of the given exemplar while encapsulating the content information of the source sentence.

GraphCheck: Breaking Long-Term Text Barriers with Extracted Knowledge Graph-Powered Fact-Checking (2025.acl-long)

Copied to clipboard

Yingjian Chen, Haoran Liu, Yinhong Liu, Jinxiang Xie, Rui Yang, Han Yuan, Yanran Fu, Peng Yuan Zhou, Qingyu Chen, James Caverlee, Irene Li

Structured Data in NLP

Challenge:	Existing fact-checking methods that use large language models often generate subtle factual errors.
Approach:	They propose a fact-checking framework that uses extracted knowledge graphs to enhance text representation.
Outcome:	GraphCheck outperforms existing specialized fact-checkers on seven benchmarks spanning general and medical domains . Graph Neural Networks process extracted knowledge graphs as a soft prompt, enabling efficient fact- checking in a single inference call.

Improving Zero and Few-Shot Abstractive Summarization with Intermediate Fine-tuning and Data Augmentation (2021.naacl-main)

Copied to clipboard

Alexander Fabbri, Simeng Han, Haoyuan Li, Haoran Li, Marjan Ghazvininejad, Shafiq Joty, Dragomir Radev, Yashar Mehdad

MUR: Momentum Uncertainty guided Reasoning for Large Language Models (2026.acl-long)

Challenge:	Recent advances in summarization are driven by the availability of large datasets such as the CNN-DailyMail corpus and the New York Times corpus.
Approach:	They propose a method for fine-tuning pretrained models for summarization in unsupervised manner . they use Wikipedia data to produce pseudo-summaries which contain characteristics of target dataset .
Outcome:	The proposed method achieves state-of-the-art, zero-shot abstractive summarization performance on CNN-DailyMail dataset and compares with other methods on other datasets.

Copied to clipboard

Hang Yan, Fangzhi Xu, Rongman Xu, Yifei Li, Jian Zhang, Haoran Luo, Xiaobao Wu, Anh Tuan Luu, Haiteng Zhao, Qika Lin, Jun Liu

A Frustratingly Simple Decoding Method for Neural Text Generation (2024.lrec-main)

Challenge:	Existing methods for optimizing reasoning quality are limited by overthinking.
Approach:	They propose a method that allocates thinking budgets to critical reasoning steps by tracking and aggregating step-wise uncertainty over time.
Outcome:	The proposed method reduces computation by over 45% on average while improving accuracy by 0.33–3.46%.

Copied to clipboard

Haoran Yang, Deng Cai, Huayang Li, Wei Bi, Wai Lam, Shuming Shi

Decoding LLM Personality Measurement: Forced-Choice vs. Likert (2025.findings-acl)

Challenge:	Neural text generation is notorious for repetitive loops and tedious outputs.
Approach:	They propose a method that penalizes future generation of repetitive content . they construct an anti-LM based on previously generated text .
Outcome:	The proposed method outperforms established baselines in terms of generation quality, decoding speed, and universality.

Copied to clipboard

Xiaoyu Li, Haoran Shi, Zengyi Yu, Yukun Tu, Chanjin Zheng

General Purpose Text Embeddings from Pre-trained Language Models for Scalable Inference (2020.findings-emnlp)

Challenge:	Recent research has focused on investigating the psychological characteristics of Large Language Models (LLMs), emphasizing the importance of comprehending their behavioral traits.
Approach:	They evaluated six Large Language Models: Llama-3.1-8B, GLM-4-9B, Claude-3.5-sonnet, and Deepseek-V3 and used the forced-choice test to assess their personality traits.
Outcome:	The forced-choice test is more reliable and more accurate than the likert scale and forced-CHOICE test results for LLMs' Big Five personality scores.

Copied to clipboard

Jingfei Du, Myle Ott, Haoran Li, Xing Zhou, Veselin Stoyanov

NegotiationToM: A Benchmark for Stress-testing Machine Theory of Mind on Negotiation Surrounding (2024.findings-emnlp)

Challenge:	Large pre-trained language models are currently used for many NLP tasks . however, inference for these models requires significant computational resources .
Approach:	They propose to use a shared text encoder to amortize the computational cost of inference over multiple tasks.
Outcome:	The proposed method reduces the size of the extracted representations by a factor of 16 to store them for later use.

Copied to clipboard

Chunkit Chan, Cheng Jiayang, Yauwai Yim, Zheye Deng, Wei Fan, Haoran Li, Xin Liu, Hongming Zhang, Weiqi Wang, Yangqiu Song

Neural Mixed Counting Models for Dispersed Topic Discovery (2020.acl-main)

Challenge:	Theory of mind evaluations currently focus on testing models using machine-generated data or game settings prone to shortcuts and spurious correlations.
Approach:	They propose a benchmark to stress-test machine ToM in real-world negotiation surrounding covered multi-dimensional mental states.
Outcome:	The proposed benchmark builds upon the Belief-Desire-Intention theory and conducts the necessary empirical experiments to evaluate large language models.

Copied to clipboard

Jiemin Wu, Yanghui Rao, Zusheng Zhang, Haoran Xie, Qing Li, Fu Lee Wang, Ziye Chen

Information Extraction & Text Mining

Challenge:	Existing methods for inference of parameter parameters are time-consuming and difficult to use.
Approach:	They propose two efficient neural mixed counting models that use the negative binomial distribution as the prior for dispersed topic discovery.
Outcome:	The proposed models outperform state-of-the-art models in terms of perplexity and topic coherence on real-world datasets.

SafeMT: Multi-turn Safety for Multimodal Language Models (2026.acl-long)

Copied to clipboard

Han Zhu, Juntao Dai, Jiaming Ji, Haoran Li, Chengkun Cai, Pengcheng Wen, Chi-Min Chan, Boyuan Chen, Yaodong Yang, Sirui Han, Yike Guo

MTOP: A Comprehensive Multilingual Task-Oriented Semantic Parsing Benchmark (2021.eacl-main)

Challenge:	Multi-turn dialogues pose a greater risk than single prompts, but existing safety benchmarks do not account for this situation.
Approach:	They propose a benchmark that features dialogues of varying lengths generated from harmful queries accompanied by images.
Outcome:	The proposed model reduces multi-turn Attack Success Rate (ASR) compared to existing guard models.

Copied to clipboard

Haoran Li, Abhinav Arora, Shuohui Chen, Anchit Gupta, Sonal Gupta, Yashar Mehdad

Self-Attention Guided Copy Mechanism for Abstractive Summarization (2020.acl-main)

Challenge:	Existing datasets for task-oriented dialog systems are limited and expensive . current models are based on the simple intent and slot detection paradigm for non-compositional queries.
Approach:	They propose to use a multilingual dataset to scale semantic parsing models to new languages . they demonstrate an average improvement of +6.3 points on Slot F1 for existing datasets .
Outcome:	The proposed model achieves an average improvement of +6.3 points on Slot F1 over existing models.

Copied to clipboard

Song Xu, Haoran Li, Peng Yuan, Youzheng Wu, Xiaodong He, Bowen Zhou

RULE: Reliable Multimodal RAG for Factuality in Medical Vision Language Models (2024.emnlp-main)

Challenge:	Abstractive summarization models have been widely used to extract words from source into summary, but how to ensure that important words in source are copied remains a challenge.
Approach:	They propose a Transformer-based model to enhance copy mechanism by identifying the importance of each source word based on the degree centrality.
Outcome:	The proposed model outperforms baseline methods on CNN/Daily Mail and Gigaword datasets.

Copied to clipboard

Peng Xia, Kangyu Zhu, Haoran Li, Hongtu Zhu, Yun Li, Gang Li, Linjun Zhang, Huaxiu Yao

BSCodec: A Band-Split Neural Codec for High-Quality Universal Audio Reconstruction (2026.findings-eacl)

Challenge:	Existing medical large vision language models often generate inaccurate and irrelevant answers that do not align with established medical facts.
Approach:	They propose a strategy for controlling factuality risk through calibrated selection of the number of retrieved contexts and a preference dataset to fine-tune the model.
Outcome:	The proposed model achieves an average improvement of 20.8% on three medical VQA datasets.

Copied to clipboard

Haoran Wang, Jiatong Shi, Jinchuan Tian, Bohan Li, Kai Yu, Shinji Watanabe

MobiLoRA: Accelerating LoRA-based LLM Inference on Mobile Devices via Context-aware KV Cache Optimization (2025.acl-long)

Challenge:	Neural audio codecs have enabled high-fidelity reconstruction of speech, music and sound . however, speech-optimized codec systems suffer degradation on music or sound if they ignore spectral differences .
Approach:	They propose a neural audio codec that splits the spectral dimension into separate bands and compresses each band independently.
Outcome:	Experimental results show that BSCodec achieves better reconstruction quality on music and sound compared to existing codecs.

Copied to clipboard

Borui Li, Yitao Wang, Haoran Ma, Ligeng Chen, Jun Xiao, Shuai Wang

Advancing Event Causality Identification via Heuristic Semantic Dependency Inquiry Network (2024.emnlp-main)

Challenge:	MobiLoRA focuses on optimizing the key-value (KV) caches due to the limited computing and memory resources of mobile devices.
Approach:	They propose to optimize the key-value caches due to limited computing resources . they propose similarity-aware delta encoding for semantic-level contexts .
Outcome:	The proposed model accelerates LoRA-based LLM inference by 57.6% on mobile devices.

Copied to clipboard

Haoran Li, Qiang Gao, Hongmei Wu, Li Huang

Knowledge Representation

Challenge:	Existing methods for ECI rely on causal features and external knowledge, but these methods fail in two dimensions: causal features between events in texts often lack explicit clues and external information may introduce bias.
Approach:	They propose a simple and effective Semantic Dependency Inquiry Network for ECI that captures semantic dependencies within the context using a unified encoder and generates a fill-in token based on comprehensive context understanding.
Outcome:	Extensive experiments show that SemDI surpasses state-of-the-art methods on three widely used benchmarks.

Internal Value Alignment in Large Language Models through Controlled Value Vector Activation (2025.acl-long)

Copied to clipboard

Haoran Jin, Meng Li, Xiting Wang, Zhihao Xu, Minlie Huang, Yantao Jia, Defu Lian

PRINCE: Prefix-Masked Decoding for Knowledge Enhanced Sequence-to-Sequence Pre-Training (2022.emnlp-main)

Challenge:	Existing LLMs do not possess consistent values, but many have been developed to align them at the behavioral level, including supervised fine-tuning (SFT) and reinforcement learning from human feedback (RLHF).
Approach:	They propose a Controlled Value Vector Activation method that directly aligns the internal values of Large Language Models by interpreting how a value is encoded in their latent representations.
Outcome:	The proposed method achieves highest success rate across 10 basic values without hurting model performance and fluency, and ensures target values even with opposite and potentially malicious input prompts.

Copied to clipboard

Song Xu, Haoran Li, Peng Yuan, Youzheng Wu, Xiaodong He

Premise-based Multimodal Reasoning: Conditional Inference on Joint Textual and Visual Clues (2022.acl-long)

Challenge:	Existing studies focus on injecting noises into the input sequence, but feasibility of injecting them into the decoding sequence remains an open question.
Approach:	They propose a pre-training paradigm that integrates knowledge-enhanced decoding with noises in the prefix to strengthen the representation learning of entities that span over multiple input tokens.
Outcome:	The proposed model achieves state-of-the-art results on two knowledge-driven data-to-text generation tasks with up to 2% BLEU gains.

Copied to clipboard

Qingxiu Dong, Ziwei Qin, Heming Xia, Tian Feng, Shoujie Tong, Haoran Meng, Lin Xu, Zhongyu Wei, Weidong Zhan, Baobao Chang, Sujian Li, Tianyu Liu, Zhifang Sui

Backdoor-Powered Prompt Injection Attacks Nullify Defense Methods (2025.findings-emnlp)

Challenge:	Existing work in vision language cross-modal reasoning uses binary or multi-choice classification based on source image and textual query.
Approach:	They propose a task where a textual premise is the background presumption on each source image.
Outcome:	The proposed task is based on a dataset of 15,360 movie screenshots and human-curated premise templates from 6 pre-defined categories.

Copied to clipboard

Yulin Chen, Haoran Li, Yuan Sui, Yangqiu Song, Bryan Hooi

ConvoSumm: Conversation Summarization Benchmark and Improved Abstractive Summarization with Argument Mining (2021.acl-long)

Challenge:	Recent studies have shown that LLMs are vulnerable to prompt injection attacks because of their instruction-following abilities and inability to distinguish the instructions in the data content.
Approach:	They propose backdoor-powered prompt injection attacks that trick LLMs into deviating from the original input instruction and executing the attackers’ target instruction.
Outcome:	The proposed attacks trick the LLMs into deviating from the input instruction and executing the attackers’ target instruction.

Copied to clipboard

Alexander Fabbri, Faiaz Rahman, Imad Rizvi, Borui Wang, Haoran Li, Yashar Mehdad, Dragomir Radev

SciCustom: A Framework for Custom Evaluation of Scientific Capabilities in Large Language Models (2026.acl-long)

Challenge:	Abstractive text summarization has primarily focused on modeling news articles . lack of standardized datasets for summarizing online conversations is a major problem .
Approach:	They propose to crowdsource four new datasets for summarizing online conversations . they incorporate argument mining through graph construction to directly model issues, viewpoints, and assertions present in a conversation.
Outcome:	The proposed models are compared against widely-used conversation summarization datasets and show comparable or improved results.

Copied to clipboard

Yiyang Gu, Junwei Yang, Junyu Luo, Ye Yuan, Bin Feng, Yingce Xia, Shufang Xie, Kaili Liu, Bohan Wu, Qi Shi, Haoran Li, Beier Xiao, Zhiping Xiao, Xiao Luo, Weizhi Zhang, Philip S. Yu, Zequn Liu, Ming Zhang

RewardDS: Privacy-Preserving Fine-Tuning for Large Language Models via Reward Driven Data Synthesis (2025.emnlp-main)

Challenge:	Existing evaluations of large language models fail to reflect fine-grained capabilities . existing benchmarks are manually curated or domain-generic, limiting scalability and alignment with real use cases.
Approach:	They propose a framework that allows custom construction of benchmarks from large-scale scientific data to evaluate application-specific scientific capabilities in LLMs.
Outcome:	The proposed framework reveals fine-grained differences in scientific capabilities that standard benchmarks overlook . it allows custom construction of benchmarks from large-scale scientific data to evaluate application-specific capabilities in LLMs.

Copied to clipboard

Jianwei Wang, Chengming Shi, Junyao Yang, Haoran Li, Qianli Ma, Huiping Zhuang, Cen Chen, Ziqian Zeng

Emerging Cross-lingual Structure in Pretrained Language Models (2020.acl-main)

Challenge:	Existing solutions to fine-tune large language models for domain-specific tasks are ineffective in addressing privacy concerns.
Approach:	They propose a privacy-preserving framework that fine-tunes a reward proxy model and uses reward signals to guide the synthetic data generation.
Outcome:	The proposed framework fine-tunes a reward proxy model and uses reward signals to guide the synthetic data generation.

Copied to clipboard

Alexis Conneau, Shijie Wu, Haoran Li, Luke Zettlemoyer, Veselin Stoyanov

InCharacter: Evaluating Personality Fidelity in Role-Playing Agents through Psychological Interviews (2024.acl-long)

Challenge:	Recent work has shown that multilingual pretraining works, but is unable to measure these effects.
Approach:	They propose to use multilingual masked language modeling to train a model on concatenated text from multiple languages to find universal latent symmetries in embedding spaces.
Outcome:	The proposed models can be trained on concatenated text from multiple languages without shared vocabulary or domain similarity.

Copied to clipboard

Xintao Wang, Yunze Xiao, Jen-tse Huang, Siyu Yuan, Rui Xu, Haoran Guo, Quan Tu, Yaying Fei, Ziang Leng, Wei Wang, Jiangjie Chen, Cheng Li, Yanghua Xiao

Challenge:	Existing methods focus on knowledge and linguistic patterns of characters.
Approach:	They propose to evaluate character fidelity of role-playing agents with psychological scales . they propose to use psychological scale to measure personality traits of RPAs based on personality traits.
Outcome:	The proposed model reproduces character fidelity with psychological scales and shows that it is effective in measuring personality traits.

Non-Autoregressive Machine Translation as Constrained HMM (2024.findings-acl)

Copied to clipboard

Haoran Li, Zhanming Jie, Wei Lu

VP-MEL: Visual Prompts Guided Multimodal Entity Linking (2025.findings-acl)

Challenge:	Autoregressive (AR) models have some drawbacks due to slow inference speed and label bias due to local normalization.
Approach:	They propose to use a left-to-right Hidden Markov Model (HMM) to control label bias in non-autoregressive translation (NAT) They propose a bi-directional HMM, which can regularize each other's biases via shared parameters.
Outcome:	The proposed models can achieve comparable performance to autoregressive Transformers using various decoding methods.

Copied to clipboard

Hongze Mi, Jinyuan Li, Zhangxuying Zhangxuying, Haoran Cheng, Jiahao Wang, Di Sun, Gang Pan

Persona-E²: A Human-Grounded Dataset for Personality-Shaped Emotional Responses to Textual Events (2026.acl-long)

Challenge:	Existing methods for multimodal entity linking rely on mention words as retrieval cues, which limits their ability to effectively utilize information from both images and text.
Approach:	They propose a visual prompt-guided multimodal entity linking task for a text-image pair . they propose VPWiki to facilitate this task and a framework to capture latent information.
Outcome:	The proposed framework outperforms baseline methods on a VPWiki dataset.

Copied to clipboard

Yuqin Yang, Haowu Zhou, Haoran Tu, Zhiwen Hui, Shiqi Yan, HaoYang Li, Dong She, Xianrong Yao, Yang Gao, Zhanpeng Jin

Challenge:	A critical bottleneck is the lack of ground-truth human data to link personality traits to emotional shifts.
Approach:	They propose a large-scale dataset to capture reader-based emotional variations across news, social media, and life narratives.
Outcome:	The proposed model captures reader-based emotional variations across news, social media, and life narratives.

MIO: A Foundation Model on Multimodal Tokens (2025.emnlp-main)

Copied to clipboard

Zekun Moore Wang, King Zhu, Chunpu Xu, Wangchunshu Zhou, Jiaheng Liu, Yibo Zhang, Jessie Wang, Ning Shi, Siyu Li, Yizhi Li, Haoran Que, Zhaoxiang Zhang, Yuanxing Zhang, Ge Zhang, Ke Xu, Jie Fu, Wenhao Huang

Investigating Crowdsourcing Protocols for Evaluating the Factual Consistency of Summaries (2022.naacl-main)

Challenge:	Existing models lack multimodal understanding capabilities, resulting in closed-source model that does not support multimodal interleaved sequences.
Approach:	They propose a foundation model built on multimodal tokens capable of understanding and generating speech, text, images, and videos in an end-to-end, autoregressive manner.
Outcome:	The proposed model is able to understand speech, text, images, and videos in an end-to-end, autoregressive manner.

Copied to clipboard

Xiangru Tang, Alexander Fabbri, Haoran Li, Ziming Mao, Griffin Adams, Borui Wang, Asli Celikyilmaz, Yashar Mehdad, Dragomir Radev

Information Extraction & Text Mining

Challenge:	Existing pre-trained summarization models produce text that is factually inconsistent with the input.
Approach:	They present a scale-based scale for Likert rating and a scoring algorithm for Best-Worst Scaling to improve crowdsourcing reliability.
Outcome:	The proposed model is more reliable than existing models on two news summarization datasets.

CONFIT: Toward Faithful Dialogue Summarization with Linguistically-Informed Contrastive Fine-tuning (2022.naacl-main)

Copied to clipboard

Xiangru Tang, Arjun Nair, Borui Wang, Bingyao Wang, Jai Desai, Aaron Wade, Haoran Li, Asli Celikyilmaz, Yashar Mehdad, Dragomir Radev

Summarization

Challenge:	Factual inconsistencies in generated summaries severely limit the practical applications of abstractive dialogue summarization.
Approach:	They propose a typology of factual errors to better understand hallucinations generated by current models and a contrastive fine-tuning strategy to improve the factual consistency and overall quality of summaries.
Outcome:	The proposed model significantly reduces all kinds of factual errors on both SAMSum dialogue summarization and AMI meeting summarizing datasets.

GrandGuard: Taxonomy, Benchmark, and Safeguards for Elderly-Chatbot Interaction Safety (2026.findings-acl)

Copied to clipboard

Changxuan Fan, Xi Yang, Yueyuan Zheng, Bin Zhou, Yuanping Wang, Wenbin Hu, Huihao Jing, Ki Sen Hung, Dazhao Du, Haoran Li, Janet Hui-wen Hsiao, Yangqiu Song

On the Faithfulness for E-commerce Product Summarization (2020.coling-main)

Challenge:	a survey of older adults shows that many LLMs mishandle elderly-specific contextual risks.
Approach:	They propose a framework to assess elderly-specific contextual risks in LLM interactions . they use a taxonomy to identify 50 fine-grained risk types across mental well-being, financial, medical, toxicity, and privacy domains .
Outcome:	a new framework assesses elderly-specific contextual risks in LLM interactions . it achieves 96.2% and 90.9% unsafe-prompt detection accuracy, respectively .

Copied to clipboard

Peng Yuan, Haoran Li, Song Xu, Youzheng Wu, Xiaodong He, Bowen Zhou

COIG-P: A High-Quality and Large-Scale Chinese Preference Dataset for Alignment with Human Values (2026.findings-eacl)

Challenge:	e-commerce product summarization requires consistency between product attributes and summary . inconsistent product summaries can mislead users and decrease public credibility .
Approach:	They propose a model to generate e-commerce product summaries with product attributes . they encode product attribute table and constrain attribute words to be presented only through copying .
Outcome:	The proposed model significantly improves the faithfulness of e-commerce product summarization tasks.

Copied to clipboard

Siwei Wu, JinCheng Ren, Xeron Du, Shuyue Guo, Xingwei Qu, Yiming Liang, Jie Liu, Yunwen Li, Tyler Loakman, Tianyu Zheng, Boyu Feng, Huaqing Yuan, Zili Wang, Jiaheng Liu, Wenhao Huang, Chenglin Cai, Haoran Que, Jian Yang, Yuelin Bai, Zekun Moore Wang, Zhouliang Yu, Qunshu Lin, Ding Pan, Yuchen Eleanor Jiang, Tiannan Wang, Wangchunshu Zhou, Shenzhi Wang, Xingyuan Bu, Minghao Liu, Guoyin Wang, Ge Zhang, Chenghua Lin

BrowseComp-Plus: A Fair and Disentangled Evaluation Benchmark for Deep Search Agents (2026.acl-long)

Challenge:	Existing Chinese preference datasets suffer from limited scale, restricted domain coverage, and insufficiently rigorous data validation.
Approach:	They propose an LLM-based data annotation pipeline with no human intervention to annotate Chinese preference datasets.
Outcome:	The proposed pipeline outperforms existing Chinese preference datasets on AlignBench and Chinese Reward Benchmark.

Copied to clipboard

Zijian Chen, Xueguang Ma, Shengyao Zhuang, Ping Nie, Kai Zou, Sahel Sharifymoghaddam, Andrew Liu, Joshua Green, Kshama Patel, Ruoxi Meng, Mingyi Su, Yanxi Li, Haoran Hong, Xinyu Shi, Xuye Liu, Hosna Oyarhoseini, Nandan Thakur, Crystina Zhang, Luyu Gao, Wenhu Chen, Jimmy Lin

InteGround: On the Evaluation of Verification and Retrieval Planning in Integrative Grounding (2025.findings-emnlp)

Challenge:	Existing benchmarks for deep search agents rely on blackbox web search APIs . dynamic and opaque web APIs hinder reproducibility and fair comparisons - authors .
Approach:	They propose a benchmark that employs a fixed corpus for controlled retrieval for deep search agents.
Outcome:	The new benchmark shows that agents that combine large language models with retrieval tools excel at complex, reasoning-intensive queries.

Copied to clipboard

Cheng Jiayang, Qianqian Zhuang, Haoran Li, Chunkit Chan, Xin Liu, Lin Qiu, Yangqiu Song

Information Retrieval

Challenge:	Existing grounding approaches work well for simple queries, but many real-world information needs require synthesizing multiple pieces of evidence.
Approach:	They introduce "integrative grounding" to evaluate the ability to ground large language models in external knowledge sources.
Outcome:	The proposed approach is robust to redundant evidence, but rationalizes using internal knowledge when information is incomplete.

Evidence Retrieval is almost All You Need for Fact Verification (2024.findings-acl)

Copied to clipboard

Liwen Zheng, Chaozhuo Li, Xi Zhang, Yu-Ming Shang, Feiran Huang, Haoran Jia

MCIP: Protecting MCP Safety via Model Contextual Integrity Protocol (2025.emnlp-main)

Challenge:	Existing evidence retrieval methods adopt a trivial retrieval strategy, resulting in task-irrelevant evidence and undesirable performance.
Approach:	They propose a framework for evidence retrieval and joint fact verification that integrates two modules.
Outcome:	The proposed framework improves evidence retrieval and claims verification on a FEVER dataset.

Copied to clipboard

Huihao Jing, Haoran Li, Wenbin Hu, Qi Hu, Xu Heli, Tianshu Chu, Peizhao Hu, Yangqiu Song

Mathematical Proof as a Litmus Test: Revealing Failure Modes of Advanced Large Reasoning Models (2026.acl-long)

Challenge:	Model Context Protocol (MCP) introduces an easy-to-use ecosystem for users and developers, but it also brings underexplored safety risks.
Approach:	They propose a framework that addresses the missing safety mechanisms in MCP and a taxonomy that captures diverse range of unsafe behaviors observed in MMP scenarios.
Outcome:	The proposed framework improves safety performance on state-of-the-art LLMs by capturing unsafe behaviors and analyzing the results.

Copied to clipboard

Dadi Guo, Jiayu Liu, Zhiyuan Fan, Zhitao He, Haoran Li, Yuxin Li, Yumeng Wang, Yi R. Fung

Reasoning

Challenge:	Large reasoning models have demonstrated remarkable mathematical problem-solving abilities, but their true reasoning shortcomings are often hidden.
Approach:	They propose to leverage the rigor and methodological complexity of mathematical proofs as a diagnostic tool to expose hidden failures.
Outcome:	The proposed model evaluation exploits the rigor and complexity of proof problems to uncover 10 fine-grained errors.

SciMMIR: Benchmarking Scientific Multi-modal Information Retrieval (2024.findings-acl)

Copied to clipboard

Siwei Wu, Yizhi Li, Kang Zhu, Ge Zhang, Yiming Liang, Kaijing Ma, Chenghao Xiao, Haoran Zhang, Bohao Yang, Wenhu Chen, Wenhao Huang, Noura Al Moubayed, Jie Fu, Chenghua Lin

Ensure the Correctness of the Summary: Incorporate Entailment Knowledge into Abstractive Sentence Summarization (C18-1)

Challenge:	Multi-modal information retrieval (MMIR) is a rapidly evolving field . current benchmarks for image-text pairings overlook the scientific domain .
Approach:	They develop a scientific domain-specific MMIR benchmark to evaluate image-text pairings using open-access research paper corpora.
Outcome:	The proposed benchmarks are based on 530K image-text pairs extracted from scientific documents with detailed captions.

Copied to clipboard

Haoran Li, Junnan Zhu, Jiajun Zhang, Chengqing Zong

Summarization

Challenge:	Existing approaches focus on improving the informativeness of the summary, but ignore the correctness.
Approach:	They propose an entailment-aware encoder and an aML-based decoder to improve the correctness of the sentence summarization task.
Outcome:	The proposed model outperforms baselines on informativeness and correctness.

Multimodal Joint Attribute Prediction and Value Extraction for E-commerce Product (2020.emnlp-main)

Copied to clipboard

Tiangang Zhu, Yue Wang, Haoran Li, Youzheng Wu, Xiaodong He, Bowen Zhou

A Data-Centric Framework for Composable NLP Workflows (2020.emnlp-demos)

Challenge:	In the real world, product attribute values are incomplete and vary over time, which hinders practical applications.
Approach:	They propose a multimodal method to jointly predict product attributes and extract values from product images using multimodal product information.
Outcome:	The proposed method can predict product attributes and extract values from product images with the help of product images.

Copied to clipboard

Zhengzhong Liu, Guanxiong Ding, Avinash Bukkittu, Mansi Gupta, Pengzhi Gao, Atif Ahmed, Shikun Zhang, Xin Gao, Swapnil Singhavi, Linwei Li, Wei Wei, Zecong Hu, Haoran Shi, Xiaodan Liang, Teruko Mitamura, Eric Xing, Zhiting Hu

Challenge:	Empirical natural language processing (NLP) systems involve interoperation among multiple components . a wealth of NLP toolkits exist ( 4), such as spaCy, DKPro, CoreNLP.
Approach:	They propose a unified open-source framework that supports fast development of NLP workflows . framework includes processors for NLP tasks, visualization, and annotation .
Outcome:	The framework offers processors for NLP tasks, visualization, and annotation, and is extensible . it is delivered through two modularized yet integratable open-source projects, Forte and Stave .

AnswerSumm: A Manually-Curated Dataset and Pipeline for Answer Summarization (2022.naacl-main)

Copied to clipboard

Alexander Fabbri, Xiaojian Wu, Srini Iyer, Haoran Li, Mona Diab

Summarization

Challenge:	Community Question Answering (CQA) fora lack a dataset to produce answer summarizations . a novel dataset of 4,631 CQA threads is used to generate answer summaries .
Approach:	They propose a dataset of 4,631 CQA threads for answer summarization curated by professional linguists.
Outcome:	The proposed approach boosts summarization performance according to automatic evaluation.

Bridging the Gap between Pre-Training and Fine-Tuning for Commonsense Generation (2023.findings-eacl)

Copied to clipboard

Haoran Yang, Yan Wang, Piji Li, Wei Bi, Wai Lam, Chen Xu

ContextLens: Modeling Imperfect Privacy and Safety Context for Legal Compliance (2026.acl-long)

Challenge:	Existing methods focusing on this task usually concatenate the concatened concepts words as the inputs of a pre-trained language model (PLM) however, in pre-training, the input is often corrupted sentences with correct word order.
Approach:	They propose a two-stage framework to improve the ability of pre-trained language models to deal with masked sentences with incorrect word order and a special token to make the input distribution more similar to the one used in pre-training.
Outcome:	The proposed method is able to generate a sentence containing all given concepts and correctly describe the relations between concepts.

Copied to clipboard

Haoran Li, Yulin Chen, Huihao Jing, Wenbin Hu, Tsz Ho Li, Chanhou Lou, Hong Ting Tsang, Sirui Han, Yangqiu Song

Multimodal Sentence Summarization via Multimodal Selective Encoding (2020.coling-main)

Challenge:	Existing approaches to contextualize safety and privacy assessments assume the availability of complete and clear context, whereas real-world contexts tend to be ambiguous and incomplete.
Approach:	They propose a semi-rule-based framework that leverages large language models to ground the input context in the legal domain and explicitly identify both known and unknown factors for legal compliance.
Outcome:	The proposed framework can significantly improve existing baselines without training and can identify the ambiguous and missing factors.

Copied to clipboard

Haoran Li, Junnan Zhu, Jiajun Zhang, Xiaodong He, Chengqing Zong

Learn to Copy from the Copying History: Correlational Copy Network for Abstractive Summarization (2021.emnlp-main)

Challenge:	Existing methods for generating summary from text and image ignore that the image can improve the ability of the encoder to identify highlights of a news event or document.
Approach:	They propose a multimodal selective gate network that takes reciprocal relationships between textual and multi-level visual features into account to select highlights of the event.
Outcome:	The proposed model can generate summary for a given sentence-image pair using visual signals . it can also capture highlights embedded in the image more accurately, the authors show .

Copied to clipboard

Haoran Li, Song Xu, Peng Yuan, Yujia Wang, Youzheng Wu, Xiaodong He, Bowen Zhou

OmniCompliance-100K: A Multi-Domain, Rule-Grounded, Real-World Safety Compliance Dataset (2026.findings-acl)

Challenge:	Existing methods for abstractive summarization use encoder-decoder attention, but this leads to incomplete copying.
Approach:	They propose a copying scheme that takes advantage of prior copying distributions and explicitly encourages the model to copy the input word that is relevant to the previously copied one.
Outcome:	The proposed scheme achieves state-of-the-art on summarization benchmarks . it takes advantage of prior copying distributions and explicitly encourages copying .

Copied to clipboard

Wenbin Hu, Huihao Jing, Haochen Shi, Changxuan Fan, Haoran Li, Yangqiu Song

Narrowing the Gap between Zero- and Few-shot Machine Translation by Matching Styles (2024.findings-naacl)

Challenge:	Existing LLM safety datasets rely on ad-hoc taxonomies and lack rule-grounded, real-world cases.
Approach:	They construct a rule-grounded, real-world case dataset OmniCompliance-100K from a compliance perspective using a powerful web-searching agent.
Outcome:	The proposed dataset spans 74 regulations and policies across a wide range of domains including security and privacy regulations, content safety and user data privacy policies, financial security requirements, medical device risk management standards, educational integrity guidelines, and protections of fundamental human rights.

Copied to clipboard

Weiting Tan, Haoran Xu, Lingfeng Shen, Shuyue Stella Li, Kenton Murray, Philipp Koehn, Benjamin Van Durme, Yunmo Chen

Privacy Checklist: Privacy Violation Detection Grounding on Contextual Integrity Theory (2025.naacl-long)

Challenge:	Recent work shows that large language models can generalize to machine translation using zero-shot examples with in-context learning.
Approach:	They investigate the factors contributing to this gap by matching the writing styles of the target corpus.
Outcome:	The proposed methods can be enhanced without the need for parallel demonstration examples.

Copied to clipboard

Haoran Li, Wei Fan, Yulin Chen, Cheng Jiayang, Tianshu Chu, Xuebing Zhou, Peizhao Hu, Yangqiu Song

Upsample or Upweight? Balanced Training on Heavily Imbalanced Datasets (2025.naacl-long)

Challenge:	Existing privacy studies focus on sub-fields, but they focus on a few sub-domains.
Approach:	They propose to use the Health Insurance Portability and Accountability Act of 1996 as an example to develop a checklist that covers social identities, private attributes, and existing privacy regulations.
Outcome:	The proposed checklist covers social identities, private attributes, and existing privacy regulations.

Copied to clipboard

Tianjian Li, Haoran Xu, Weiting Tan, Kenton Murray, Daniel Khashabi

GoldCoin: Grounding Large Language Models in Privacy Laws via Contextual Integrity Theory (2024.emnlp-main)

Challenge:	a lack of data across domains creates significant imbalances in training data sizes . a recent study shows that temperature sampling and scaling are equivalent but differ under stochastic gradient descent due to differences in gradient variance.
Approach:	They propose a method that upsamples low-resource languages and upweights their loss functions to address this disparity.
Outcome:	The proposed method competes effectively with existing data re-weighting techniques while offering computational efficiency.

Copied to clipboard

Wei Fan, Haoran Li, Zheye Deng, Weiqi Wang, Yangqiu Song

Multi-step Jailbreaking Privacy Attacks on ChatGPT (2023.findings-emnlp)

Challenge:	Existing research studies privacy by exploring various privacy attacks, defenses, and evaluations within narrowly predefined patterns.
Approach:	They propose a framework that leverages the theory of contextual integrity as a bridge to help LLMs understand the complex contexts for judicial assessing privacy violations.
Outcome:	The proposed framework bridges the theory of contextual integrity as a bridge, creating numerous synthetic scenarios grounded in relevant privacy statutes (e.g., HIPAA).

Copied to clipboard

Haoran Li, Dadi Guo, Wei Fan, Mingshi Xu, Jie Huang, Fanpu Meng, Yangqiu Song

ToMELP: A Theory-of-Mind Benchmark for Route-Controlled Persuasion under the Elaboration Likelihood Model (2026.findings-acl)

Challenge:	With the rapid evolution of large language models (LLMs), many downstream NLP tasks can be well solved given appropriate prompts.
Approach:	They propose to integrate ChatGPT and Bing GPT3 into their applications to create a set of LLMs that can be used to generate NLP tasks with appropriate prompts.
Outcome:	The proposed models can be zero-shot or few-shot learners to solve specified tasks and can even be zero or few shot learners.

Copied to clipboard

Ruirui Wang, Haoran Zhang, Tian Lan, Zehua Duo, Jiang Li, Guanglai Gao, Xiangdong Su

Linguistics & Cognitive NLP

Challenge:	Theory of Mind (ToM) is widely regarded as central to effective persuasion, yet existing evaluations fail to capture the infer–apply loop that arises in real-world dialogue.
Approach:	They propose a benchmark that conditions on the audience persona p and the Elaboration Likelihood Model (ELM) route r within persuasive conversations.
Outcome:	The proposed model can model the interlocutor's mental states over multiple turns and adapt strategy and tone accordingly.

Spelling Error Correction with Soft-Masked BERT (2020.acl-main)

Copied to clipboard

Shaohua Zhang, Haoran Huang, Jicong Liu, Hang Li

Into the Gray Zone: Domain Contexts Can Blur LLM Safety Boundaries (2026.acl-long)

Challenge:	Experimental results show that the proposed method is significantly better than the baselines including the one solely based on BERT.
Approach:	They propose a neural architecture which uses a network for error detection and a system for error correction based on BERT, with the latter connected to the other using what they call soft-masking technique.
Outcome:	The proposed method performs better than baselines including the one solely based on BERT, and is general and may be employed in other language detection-correction problems.

Copied to clipboard

Ki Sen Hung, Xi Yang, Chang Liu, Haoran Li, Kejiang Chen, Changxuan Fan, Tsun On Kwok, Weiming Zhang, Xiaomeng Li, Yangqiu Song

STRUDEL: Structured Dialogue Summarization for Dialogue Comprehension (2022.emnlp-main)

Challenge:	a goal of LLM alignment is to balance usefulness with harmlessness, but this conflictes when knowledge serves both legitimate and malicious purposes.
Approach:	They propose a framework that combines safety-research contexts with adversarial interactions to exploit a vulnerability in Jargon queries.
Outcome:	a framework outperforms existing methods in analyzing Jargon queries, a study shows . it achieves 93% of attacks across seven models, while remaining useful, the authors say .

Copied to clipboard

Borui Wang, Chengcheng Feng, Arjun Nair, Madelyn Mao, Jai Desai, Asli Celikyilmaz, Haoran Li, Yashar Mehdad, Dragomir Radev

Natural Language Interfaces

Challenge:	Abstractive dialogue summarization is an important standalone task in natural language processing, but no previous work has explored whether it can be used to boost an NLP system's performance on other important dialogue comprehension tasks.
Approach:	They propose a novel type of dialogue summarization task that decomposes and imitates the hierarchical, systematic and structured mental process that human beings usually go through when understanding and analyzing dialogues.
Outcome:	The proposed model improves the performance of transformer encoder language models on two important dialogue comprehension tasks.

PrivLM-Bench: A Multi-level Privacy Evaluation Benchmark for Language Models (2024.acl-long)

Copied to clipboard

Haoran Li, Dadi Guo, Donghao Li, Wei Fan, Qi Hu, Xin Liu, Chunkit Chan, Duanyi Yao, Yuan Yao, Yangqiu Song

Jailbreaking Large Language Models with Morality Attacks (2026.findings-acl)

Challenge:	generative large language models (LLMs) exhibit surprising capability and integrate previous tasks into a unified text generation formulation.
Approach:	They propose a privacy evaluation benchmark to quantify the privacy leakage of language models.
Outcome:	The proposed benchmark compares PPLMs with different privacy implementations to find out how privacy leakage is handled.

Copied to clipboard

Ying Su, Zheng Mingen, Weili Diao, Haoran Li

Activation-Guided Local Editing for Jailbreaking Attacks (2026.acl-long)

Challenge:	Pluralism alignment is the goal of creating AI that can coexist with and serve morally multifaceted humanity.
Approach:	They propose to use jailbreak attacks to manipulate LLMs’ judgment over pluralistic values by using a morality dataset with 10.4K instances.
Outcome:	The proposed method exploits the persuasion abilities of LLMs to produce moral content over pluralistic values.

Copied to clipboard

Jiecong Wang, Haoran Li, Hao Peng, Ziqian Zeng, Zihao Wang, Haohua Du, Zhengtao Yu

Robustness in NLP

Challenge:	Existing methods for jailbreaking Large Language Models (LLMs) are limited and produce incoherent or unreadable inputs.
Approach:	They propose a two-stage framework that performs a one-shot, scenario-based generation of context and rephrases the original malicious query to obscure its harmful intent.
Outcome:	The proposed framework achieves state-of-the-art Attack Success Rate, with gains of up to 37.74% over the strongest baseline, and excellent transferability to black-box and large-scale models.

A.S.E: A Repository-Level Benchmark for Evaluating Security in AI-Generated Code (2026.findings-acl)

Copied to clipboard

Keke Lian, Wang Bin, Lei Zhang, Libo Chen, Junjie Wang, Ziming Zhao, Yujiu Yang, Miaoqian Lin, Haotong Duan, Haoran Zhao, Shuang Liao, Mingda Guo, Quan Jiazheng, Yilu Zhong, Chenhao He, Chen Zichuan, Jie Wu, Haoling Li, Zhaoxuan Li, Jiongchi Yu, Hui LI, Dong Zhang

Programming Languages in NLP

Challenge:	Existing security evaluation benchmarks lack relevance to real-world AI programming tasks . current LLMs struggle with secure coding, research shows .
Approach:	They propose a repository-level evaluation benchmark to assess security of AI-generated code.
Outcome:	The proposed framework mirrors real-world AI programming tasks and offers valuable insights into the state of AI code generation.

RAP: Efficient Text-Video Retrieval with Sparse-and-Correlated Adapter (2024.findings-acl)

Copied to clipboard

Meng Cao, Haoran Tang, Jinfa Huang, Peng Jin, Can Zhang, Ruyang Liu, Long Chen, Xiaodan Liang, Li Yuan, Ge Li

Green & Sustainable NLP

Challenge:	Text-Video Retrieval (TVR) aims to align relevant video content with natural language queries.
Approach:	They propose to conduct efficient text-video Retrieval with a salient-and-correlated AdaPter . they propose a low-rank modulation module to refine per-image features from frozen CLIP backbone .
Outcome:	Experiments on four TVR datasets show that the proposed method performs better than other methods.

MSMO: Multimodal Summarization with Multimodal Output (D18-1)

Copied to clipboard

Junnan Zhu, Haoran Li, Tianshang Liu, Yu Zhou, Jiajun Zhang, Chengqing Zong

PrivaCI-Bench: Evaluating Privacy with Contextual Integrity and Legal Compliance (2025.acl-long)

Challenge:	Existing studies show that multimodal summarization can improve user satisfaction for informativeness of summaries by using information in visual modality.
Approach:	They propose a task to generate text and select the most relevant image from the multimodal input and a novel multimodal automatic evaluation method to evaluate multimodal outputs.
Outcome:	The proposed method improves user satisfaction by 12.4% compared to the current system .

Copied to clipboard

Haoran Li, Wenbin Hu, Huihao Jing, Yulin Chen, Qi Hu, Sirui Han, Tianshu Chu, Peizhao Hu, Yangqiu Song