Papers by Jiajun Li
Muse: Towards Reproducible Long-Form Song Generation with Fine-Grained Style Control (2026.findings-acl)
Copied to clipboard
Changhao Jiang, Jiahao Chen, Zhenghao Xiang, Zhixiong Yang, Hanchen Wang, Jiabao Zhuang, Xinmeng Che, Jiajun Sun, Hui Li, Yifei Cao, Shihan Dou, Ming Zhang, Junjie Ye, Tao Ji, Tao Gui, Qi Zhang, Xuanjing Huang
| Challenge: | Recent commercial systems such as Suno demonstrate strong capabilities in long-form song generation, but academic research remains non-reproducible due to the lack of publicly available training data. |
| Approach: | They propose a system for long-form song generation with fine-grained style conditioning that includes a licensed synthetic dataset and a song generation model, Muse. |
| Outcome: | The proposed system achieves competitive performance on phoneme error rate, text–music style similarity, and audio aesthetic quality while enabling controllable segment-level generation across different musical structures. |
DentalGPT: Incentivizing Multimodal Reasoning in Dentistry (2026.findings-acl)
Copied to clipboard
Zhenyang Cai, Jiaming Zhang, Junjie Zhao, Ziyi Zeng, Yanchao Li, Liang Jingyi, Junying Chen, Yunjin Yang, Jiajun You, Shuzhi Deng, null Xieruiqiii, Yuanting Chen, Xiangyi Feng, Jianquan Li, Liangyi Chen, Junwen Wang, Shan Jiang, Benyou Wang
| Challenge: | Current multimodal large language models (MLLMs) show limited understanding of dental images. |
| Approach: | They propose a dental-specialized multimodal large language model trained via staged multimodal alignment and reinforcement learning. |
| Outcome: | The proposed model outperforms state-of-the-art models on disease classification and dental VQA tasks. |
AgentGym2: Benchmarking Large Language Model Agents in De-Idealized Real-World Environments (2026.acl-long)
Copied to clipboard
Zhiheng Xi, Dingwen Yang, Jiaqi Liu, Jixuan Huang, Honglin Guo, Baodai Huang, Tinggang Chen, Qi Zhang, Zhonghang Lu, Chenyu Liu, Jiajun Sun, Jiazheng Zhang, Dingwei Zhu, Xin Guo, Junzhe Wang, Zhihao Zhang, Yuming Yang, Junjie Ye, Minghe Gao, Dongrui Liu, Jiaming Ji, Guohao Li, Tao Gui, Qi Zhang, Xuanjing Huang
| Challenge: | Existing benchmarks evaluate agents in simplified, idealized settings, relying on pre-packaged tool interfaces, overlooking critical steps, and assume inputs are clean and fully specified. |
| Approach: | They propose a framework that evaluates language agents in simplified, idealized settings . they show that even SOTA systems like Gemini and GPT-5 struggle on AgentGym2 . |
| Outcome: | Experiments on 15 proprietary and open-source models show that even SOTA systems like Gemini and GPT-5 struggle on AgentGym2 . |
Formality is Favored: Unraveling the Learning Preferences of Large Language Models on Data with Conflicting Knowledge (2024.emnlp-main)
Copied to clipboard
| Challenge: | Large language models have shown excellent performance on knowledge-intensive tasks, but pretraining data tends to contain misleading and conflicting information. |
| Approach: | They systematically analyze LLMs’ learning preferences for data with conflicting knowledge. |
| Outcome: | The proposed model outperforms human-level models on knowledge-intensive tasks by analyzing pretraining data. |
LoopCoder: Scaling Code Intelligence via Looped Language Models (2026.findings-acl)
Copied to clipboard
Jian Yang, Wei Zhang, Shuyue Guo, Yizhi LI, Linzheng Chai, Zhengmao Ye, Shukai Liu, Yuyang Song, Jiajun Wu, Che Liu, Tianyu Zheng, Siwei Wu, Leo L, Xudong Ma, Chuan Hao, Ran Tao, Yan Xing, Jianzhou Wang, Mingjie Tang, Aishan Liu, Zhoujun Li, Xianglong Liu, Weifeng Lv, Bryan Dai
| Challenge: | Large language models have mastered syntax-level code generation, but complex algorithmic reasoning remains a challenge. |
| Approach: | They propose a recurrent inductive bias that aligns with the recursive nature of programming logic. |
| Outcome: | The proposed model achieves comparable performance to standard dense models with more parameters. |
On the Consistency of Commonsense in Large Language Models (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing evaluations of commonsense for large language models focus on downstream knowledge tasks, failing to probe whether LLMs truly understand and utilize knowledge or merely memorize it. |
| Approach: | They propose to automatically construct a large benchmark named CoCo which measures LLMs’ knowledge memorization, comprehension, and application and examines the consistency between these tasks. |
| Outcome: | The proposed benchmark systematically assesses LLMs’ knowledge memorization, comprehension, and application and examines the consistency between these tasks. |
Data Whisperer: Efficient Data Selection for Task-Specific LLM Fine-Tuning via Few-Shot In-Context Learning (2025.acl-long)
Copied to clipboard
Shaobo Wang, Xiangqi Jin, Ziming Wang, Jize Wang, Jiajun Zhang, Kaixin Li, Zichen Wen, Zhong Li, Conghui He, Xuming Hu, Linfeng Zhang
| Challenge: | Using fine-tuning on task-specific data is essential for large language models to be effective in specialized tasks. |
| Approach: | They propose a method that leverages few-shot in-context learning with the model to be fine-tuned. |
| Outcome: | The proposed method outperforms existing methods with a 3.1-point improvement and a 7.4 speedup on the Llama-3-8B-Instruct model using just 10% of the dataset. |
Acquisition and Application of Novel Knowledge in Large Language Models (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for constructing new datasets rely on timestamps or simple templates that do not accurately reflect the real world. |
| Approach: | They propose a knowledge dataset construction approach that simulates biological evolution using knowledge graphs to generate synthetic entities with diverse attributes. |
| Outcome: | The proposed framework outperforms knowledge augmentation methods by 3.3%-38%. |
X-Instruction: Aligning Language Model in Low-resource Languages with Self-curated Cross-lingual Instructions (2024.findings-acl)
Copied to clipboard
| Challenge: | Large language models respond well in high-resource languages but struggle in low-resourced languages. |
| Approach: | They propose a method to construct cross-lingual instruction following samples with instruction in English and response in low-resource languages. |
| Outcome: | The proposed method builds a large-scale cross-lingual instruction tuning dataset on 10 languages. |
RealChart2Code: Bridging the Gap in Real-World Chart-to-Code Generation via Multi-Task Evaluation (2026.acl-long)
Copied to clipboard
Jiajun Zhang, Yuying Li, Zhixun Li, Xingyu Guo, Jingzhuo Wu, Leqi Zheng, Yiran Yang, Jianke Zhang, Qingbin Li, Shannan Yan, Changguo Jia, Junfei Wu, Zilei Wang, Qiang Liu, Liang Wang
| Challenge: | Vision-Language Models (VLMs) have demonstrated impressive capabilities in code generation across various domains, but their ability to replicate complex, multi-panel visualizations remains largely unassessed. |
| Approach: | They propose a large-scale benchmark to evaluate chart generation from large- scale raw data and assess iterative code refinement in a multi-turn conversational setting. |
| Outcome: | The new benchmark evaluates 14 leading VLMs on real-world data and shows they struggle with complex plot structures and authentic data. |
Explicit Semantic Decomposition for Definition Generation (2020.acl-main)
Copied to clipboard
| Challenge: | Existing definition generation methods rely on decoding to extract semantic components of words. |
| Approach: | They propose a method which explicitly decomposes meaning of words into semantic components and models them with discrete latent variables for definition generation. |
| Outcome: | The proposed method outperforms existing methods on WordNet and Oxford benchmarks. |
A Data-Efficient Path to Multilingual LLMs: Language Expansion via Post-training PARAM𝛥 Integration into Upcycled MoE (2026.acl-long)
Copied to clipboard
Hao Zhou, Tianhao Li, Zhijun Wang, Shuaijie She, Linjuan Wu, Hao-Ran Wei, Baosong Yang, Jiajun Chen, Shujian Huang
| Challenge: | Large Language Models (LLMs) are expensive and require extensive Continued Pre-Training and data-intensive alignment to expand. |
| Approach: | They propose a method which upcycles a dense model into a Mixture-of-Experts architecture, allocating different experts to different languages. |
| Outcome: | Experiments show that the proposed model upcycles a dense model into a Mixture-of-Experts(MoE) architecture, allocating different experts to different languages. |
EXPLAIN: Enhancing Retrieval-Augmented Generation with Entity Summary (2025.acl-industry)
Copied to clipboard
Yaozhen Liang, Xiao Liu, Jiajun Yu, Zhouhua Fang, Qunsheng Zou, Linghan Zheng, Yong Li, Zhiwei Liu, Haishuai Wang
| Challenge: | Existing document question answering methods reduce inference costs and input tokens. |
| Approach: | They propose a retrieval-augmented generation method that automatically extracts useful entities and generates summaries from documents. |
| Outcome: | The proposed method surpasses baseline retrieval-augmented generation (RAG) and long-context question answering (LC) methods achieve higher accuracy by processing entire documents, but at the cost of increased computational Corresponding authors. |
Generating Sentences from Disentangled Syntactic and Semantic Spaces (P19-1)
Copied to clipboard
| Challenge: | Variational auto-encoders (VAEs) are widely used in natural language generation due to the regularization of the latent space. |
| Approach: | They propose to generate sentences from disentangled syntactic and semantic spaces by using the linearized tree sequence. |
| Outcome: | The proposed method achieves similar or better performance in various tasks compared with state-of-the-art models. |
TokAlign: Efficient Vocabulary Adaptation via Token Alignment (2025.acl-long)
Copied to clipboard
| Challenge: | Tokenization is a foundational step for Large Language Models (LLMs) but low compression rate of vanilla tokenizers decelerates training and inference process. |
| Approach: | They propose a method to replace the vocabulary of Large Language Models (LLMs) by learning a one-to-one mapping matrix for token IDs. |
| Outcome: | The proposed method significantly improves multilingual text compression rates and vocabulary initialization for Large Language Models. |
Are You for Real? Detecting Identity Fraud via Dialogue Interactions (D19-1)
Copied to clipboard
| Challenge: | Existing methods to detect identity fraud are prone to errors and are not based on real data. |
| Approach: | They propose to use a KG constructor and structured dialogue management to detect identity fraud in loan applications to generate questions based on personal information. |
| Outcome: | The proposed system can detect fraudsters and achieve higher recognition accuracy compared with rule-based systems. |
Incremental Learning from Scratch for Task-Oriented Dialogue Systems (P19-1)
Copied to clipboard
| Challenge: | Existing task-oriented dialogue systems cannot guarantee that all user needs are taken into account in the design phase. |
| Approach: | They propose a new incremental learning framework to design task-oriented dialogue systems without pre-defining user needs. |
| Outcome: | The proposed framework is robust to unconsidered user actions and can update itself online with less annotation cost. |
MARCH: Multi-Agent Reinforced Check for Hallucination (2026.acl-long)
Copied to clipboard
Zhuo Li, Yupeng Zhang, Pengyu Cheng, Jiajun Song, Mengyu Zhou, Hao Li, Shujie Hu, Yu Qin, null Erchao.zec, Xiaoxi Jiang, null Guanjunjiang
| Challenge: | Existing methods to detect hallucinations suffer from inherent confirmation bias, where the verifier inadvertently reproduces the errors of the original generation. |
| Approach: | They propose a framework that enforces rigorous factual alignment by leveraging deliberate *information asymmetry* by combining a pipeline of three specialized agents: a Solver, a Proposer, and a Checker. |
| Outcome: | Extensive experiments across hallucination benchmarks demonstrate that MARCH substantially reduces hallucinism rates. |
One Sentence One Model for Neural Machine Translation (L18-1)
Copied to clipboard
| Challenge: | Neural machine translation (NMT) is a new state of the art that can produce better results than traditional statistical machine translation. |
| Approach: | They propose a dynamic neural network which learns a general network as usual and fine-tunes it for each test sentence. |
| Outcome: | The proposed method improves translation performance when similar sentences are available. |
Translate the Beauty in Songs: Jointly Learning to Align Melody and Translate Lyrics (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Song translation requires both translation of lyrics and alignment of music notes . human translators of songs need to have a mastery of cultural traditions and the poetic usage of both source and target languages . |
| Approach: | They propose a model that can model lyric translation and lyrics-melody alignment . they use an encoder-decoder framework that can translate lyrics and determine number of aligned notes . |
| Outcome: | The proposed framework can translate lyrics and determine the number of aligned notes at each decoding step. |
PreAlign: Boosting Cross-Lingual Transfer by Early Establishment of Multilingual Alignment (2024.emnlp-main)
Copied to clipboard
| Challenge: | Large language models exhibit reasonable multilingual abilities, despite predominantly English-centric pretraining. |
| Approach: | They propose a framework that establishes multilingual alignment prior to language model pretraining and preserves this alignment using a code-switching strategy during pretraining. |
| Outcome: | Experiments in a synthetic English to English-Clone setting show that PreAlign outperforms standard multilingual joint training in language modeling, zero-shot cross-lingual transfer, and cross-linguistic knowledge application. |
Analyzing the Intensity of Complaints on Social Media (2022.findings-naacl)
Copied to clipboard
| Challenge: | Prior studies on identifying the existence or the type of complaints focus on building automatic classification models for identifying complaints. |
| Approach: | They propose to measure the intensity of complaints from text using Best-Worst Scaling method to estimate the popularity of posts on social media. |
| Outcome: | The proposed model can estimate the popularity of complaints on social media with best-worst scaling (BWS) method. |
Group then Scale: Dynamic Mixture-of-Experts Multilingual Language Model (2025.findings-acl)
Copied to clipboard
| Challenge: | Experimental results show that language group specialization on experts improves multilingual performance. |
| Approach: | They propose to dynamically group and scale up parameters of multilingual Large Language Models while boosting positive transfer among similar languages. |
| Outcome: | The proposed method reduces negative transfer between languages and boosts performance on 18 to 128 languages. |
CIF-Bench: A Chinese Instruction-Following Benchmark for Evaluating the Generalizability of Large Language Models (2024.findings-acl)
Copied to clipboard
Yizhi Li, Ge Zhang, Xingwei Qu, Jiali Li, Zhaoqun Li, Noah Wang, Hao Li, Ruibin Yuan, Yinghao Ma, Kai Zhang, Wangchunshu Zhou, Yiming Liang, Lei Zhang, Lei Ma, Jiajun Zhang, Zuowen Li, Wenhao Huang, Chenghua Lin, Jie Fu
| Challenge: | a recent study shows that large language models have limited generalization in low-resource languages like Chinese. |
| Approach: | They propose to evaluate the zero-shot generalizability of large language models to the Chinese language . they release only half of the dataset publicly, with the remainder kept private . |
| Outcome: | The Chinese Instruction-Following Benchmark evaluates the generalizability of LLMs to the Chinese language. |
Rethinking Document-level Neural Machine Translation (2022.findings-acl)
Copied to clipboard
| Challenge: | Neural machine translation models are weak enough for document-level translation . current models only translate sentences individually, resulting in poor document coherence . |
| Approach: | They propose to use the original Transformer model to test document-level neural machine translation . they find that the original transformer models can achieve strong results for document translation if trained properly . |
| Outcome: | The proposed model outperforms sentence-level models on nine datasets and two sentence- level datasets across six languages. |
OMIBench: Benchmarking Olympiad-Level Multi-Image Reasoning in Large Vision-Language Models (2026.acl-long)
Copied to clipboard
Qiguang Chen, Chengyu Luan, Jiajun Wu, Qiming Yu, Yi Yang, Yizhuo Li, Jingqi Tong, Xiachong Feng, Libo Qin, Wanxiang Che
| Challenge: | Existing multimodal reasoning benchmarks for large vision-language models emphasize single-image analysis and fail to exploit contextual information across multiple images. |
| Approach: | They propose a benchmark to evaluate Olympiad-level reasoning when evidence is distributed over multiple images. |
| Outcome: | The proposed model outperforms existing models on bi-image Olympiads and Gemini-3-Pro on multimodal Olympiad-level reasoning tasks. |
Measuring Meaning Composition in the Human Brain with Composition Scores from Large Language Models (2024.acl-long)
Copied to clipboard
| Challenge: | Existing computational metric to quantify extent of meaning composition is lacking . despite extensive neurolinguistic research, understanding how meaning is constructed is difficult . |
| Approach: | They propose a computational metric to quantify the degree of meaning composition . they use key-value memory interpretation of transformer feed-forward network blocks to study meaning composition. |
| Outcome: | Experimental results show that the metric correlates with brain clusters associated with word frequency, structural processing, and general sensitivity to words, suggesting multifaceted nature of meaning composition during human sentence comprehension. |
MdEval: Massively Multilingual Code Debugging (2026.findings-acl)
Copied to clipboard
Shukai Liu, Linzheng Chai, Jian Yang, Jiajun Shi, He Zhu, Liran Wang, Jin Ke, Wei Zhang, Hualei Zhu, Shuyue Guo, Tao Sun, Jiaheng Liu, Yunlong Duan, Yu Hao, Liqun Yang, Guanglin Niu, Ge Zhang, Zhoujun Li
| Challenge: | Existing benchmarks primarily focus on Python and are limited in terms of language diversity. |
| Approach: | They propose a multilingual debugging benchmark that includes 3.9K test samples of 20 programming languages and introduces the debug instruction corpora MdEval-Instruct by injecting bugs into the correct multilingual queries and solutions. |
| Outcome: | The proposed benchmark includes 3.9K test samples of 20 programming languages and covers the automated program repair task, bug localization task, and bug identification task. |
Learning Confidence for Transformer-based Neural Machine Translation (2022.acl-long)
Copied to clipboard
| Challenge: | A well-calibrated confidence estimate is not sufficient for neural machine translation (NMT) where probabilities from softmax distribution fail to describe when the model is probably mistaken. |
| Approach: | They propose an unsupervised confidence estimate learning jointly with the training of a neural machine translation model to quantify confidence. |
| Outcome: | The proposed model outperforms standard label smoothing and can predict failures in two real-world scenarios. |
A Teacher-Student Framework for Maintainable Dialog Manager (D18-1)
Copied to clipboard
| Challenge: | Reinforcement learning (RL) is an attractive solution for task-oriented dialog systems . but extending RL-based systems to handle new intents and slots requires a system redesign . |
| Approach: | They propose a teacher-student framework to extend RL-based dialog systems . they propose to specify constraints held in the new dialog manager . |
| Outcome: | The proposed framework makes no assumption about unsupported intents and slots, making it possible to improve RL-based systems incrementally. |
VehicleWorld: A Highly Integrated Multi-Device Environment for Intelligent Vehicle Interaction (2025.findings-emnlp)
Copied to clipboard
Jie Yang, Jiajun Chen, Zhangyue Yin, Shuo Chen, Yuxin Wang, Yiran Guo, Yuan Li, Yining Zheng, Xuanjing Huang, Xipeng Qiu
| Challenge: | Traditional Function Calling (FC) approaches operate statelessly, requiring multiple exploratory calls to build environmental awareness before execution, leading to inefficiency and limited error recovery. |
| Approach: | They propose a state-based function call approach that maintains explicit system state awareness and implements direct state transitions to achieve target conditions. |
| Outcome: | The proposed approach outperforms traditional function calling approaches, achieving superior execution accuracy and reduced latency. |
SAFETY-J: Evaluating Safety with Critique (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Current methods focus on binary safety classifications and lack detailed critique, limiting their utility for model improvement and user trust. |
| Approach: | They propose a bilingual generative safety evaluator for English and Chinese with critique-based judgment that utilizes a robust training dataset and augmented query-response pairs to assess safety across various scenarios comprehensively. |
| Outcome: | The proposed model improves safety evaluations by assessing the quality of critiques with minimal human intervention. |
MTAVG-Bench: A Diagnostic Benchmark for Multi-Talker Dialogue-Centric Audio-Video Generation (2026.acl-long)
Copied to clipboard
Yanghao Zhou, Haitian Li, Rexar Lin, Heyan Huang, Jinxing Zhou, Changsen Yuan, Tian Lan, Ziqin Zhou, Yudong Li, Jiajun Xu, Jingyun Liao, YiMing Cheng, Xuefeng Chen, Xian-Ling Mao, Yousheng Feng
| Challenge: | Existing evaluation benchmarks for text-to-audio-video (T2AV) generation are largely designed for human-recorded videos or single-speaker settings. |
| Approach: | They propose a failure-driven diagnostic benchmark for multi-talker dialogue-centric audio-video generation. |
| Outcome: | The benchmark evaluates multi-speaker dialogue generation at four levels: audio-visual signal fidelity, temporal attribute consistency, social interaction, and cinematic expression. |
MMAD:Multi-modal Movie Audio Description (2024.lrec-main)
Copied to clipboard
| Challenge: | Current methods of creating accessible movies rely on manual work, resulting in high costs and limited scalability. |
| Approach: | They propose a multi-modal movie audio description pipeline that generates narrations of information that is not accessible through unimodal hearing in movies. |
| Outcome: | The proposed pipeline surpasses existing baselines in performance on widely used datasets. |
LR²Bench: Evaluating Long-chain Reflective Reasoning Capabilities of Large Language Models via Constraint Satisfaction Problems (2025.findings-acl)
Copied to clipboard
| Challenge: | Recent advances in o1-like models have significantly enhanced the reasoning abilities of Large Language Models (LLMs). |
| Approach: | They propose a benchmark to evaluate the Long-chain Reflective Reasoning capabilities of Large Language Models. |
| Outcome: | The proposed benchmark evaluates the Long-chain Reflective Reasoning capabilities of Large Language Models (LLMs) it consists of 850 samples across six Constraint Satisfaction Problems (CSPs) |
Multilingual Machine Translation with Large Language Models: Empirical Results and Analysis (2024.findings-naacl)
Copied to clipboard
Wenhao Zhu, Hongyi Liu, Qingxiu Dong, Jingjing Xu, Shujian Huang, Lingpeng Kong, Jiajun Chen, Lei Li
| Challenge: | Existing studies show that large language models (LLMs) can handle multilingual machine translation (MMT) However, the multilingual translation ability of LLMs remains under-explored. |
| Approach: | They evaluate eight popular LLMs including ChatGPT and GPT-4 to determine their performance in multilingual machine translation. |
| Outcome: | The proposed model can generate moderate translation even on zero-resource languages and cross-lingual exemplars can provide better task guidance for low-resourced translation than exemplar in the same language pairs. |
EfficientLLM: Unified Pruning-Aware Pretraining for Auto-Designed Compact Language Models (2026.acl-long)
Copied to clipboard
Xingrun Xing, Zheng Liu, Shitao Xiao, Boyan Gao, Yiming Liang, Haokun Lin, Xianlin Zeng, Guoqi Li, Jiajun Zhang
| Challenge: | Large language models (LLMs) driven by scaling laws can be developed in large model sizes. |
| Approach: | They propose a pruning-aware pretraining approach that decouples LLM pruning from direct pretraining. |
| Outcome: | The proposed model outperforms pretraining models with 100M 1B parameters in commen sense benchmarks. |
Multilingual Pretraining and Instruction Tuning Improve Cross-Lingual Knowledge Alignment, But Only Shallowly (2024.naacl-long)
Copied to clipboard
| Challenge: | Current large language models show imbalance abilities in different languages . authors propose two approaches to improve cross-lingual knowledge alignment . |
| Approach: | They propose a framework to assess cross-lingual knowledge alignment of large language models . they propose multilingual pretraining and multilingual instruction tuning to address this problem . |
| Outcome: | The proposed framework assesses the cross-lingual knowledge alignment of LLMs in performance, consistency and conductivity levels. |
Interpreting and Exploiting Functional Specialization in Multi-Head Attention under Multi-task Learning (2023.emnlp-main)
Copied to clipboard
| Challenge: | Experimental results show that multi-head attention module evolves functional specialization after multi-task training. |
| Approach: | They propose a method to quantify the degree of functional specialization in multi-head attention . they propose 'multi-task training' method to increase functional specialisation and mitigate negative information transfer . |
| Outcome: | The proposed method increases functional specialization and mitigates negative information transfer in multi-task learning without adding any parameters. |
SDGO: Self-Discrimination-Guided Optimization for Consistent Safety in Large Language Models (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) excel at various tasks but are vulnerable to jailbreak attacks that induce harmful content generation. |
| Approach: | They propose a reinforcement learning framework that leverages the model’s own discrimination capabilities as a reward signal to enhance generation safety through iterative self-improvement. |
| Outcome: | The proposed framework improves model safety by iterative self-improvement without additional annotated data or external models during training phase. |
LLM-Guided Semantic-Aware Clustering for Topic Modeling (2025.acl-long)
Copied to clipboard
| Challenge: | Experimental results show that topic modeling is competitive compared to closed-source methods. |
| Approach: | They propose a semi-supervised topic modeling method that combines LLMs with clustering to improve topic generation and distribution. |
| Outcome: | The proposed method outperforms state-of-the-art methods that utilize GPT-4 on topic alignment and exhibits competitive performance compared to Neural Topic Models on topic quality. |
Improving In-context Learning of Multilingual Generative Language Models with Cross-lingual Alignment (2024.naacl-long)
Copied to clipboard
| Challenge: | Existing studies show that multilingual generative models exhibit a strong language bias toward high-resource languages. |
| Approach: | They propose a cross-lingual alignment framework exploiting pairs of translation sentences to improve cross-linguistic abilities. |
| Outcome: | The proposed framework improves cross-lingual abilities and mitigates performance gap. |
Roles of Scaling and Instruction Tuning in Language Perception: Model vs. Human Attention (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Recent large language models (LLMs) have shown strong abilities to understand natural language, but how these factors affect the models’ language perception is unclear. |
| Approach: | They compare the self-attention of several existing large language models in different sizes to assess the effect of scaling and instruction tuning on language perception. |
| Outcome: | The proposed models are closer to non-native speakers than native speakers in attention, suggesting a sub-optimal language perception of all models. |
Boosting Textural NER with Synthetic Image and Instructive Alignment (2024.findings-acl)
Copied to clipboard
| Challenge: | Named entity recognition (NER) is a key task reliant on textual data. |
| Approach: | They propose a method to transform NER into a multimodal task by using images from the internet as auxiliaries. |
| Outcome: | The proposed method surpasses all text-only baselines and improves F1 score by 1.4% to 2.3% on prominent MNER datasets. |
On the Role of Discriminative Models in Generative Relation Extraction (2026.acl-long)
Copied to clipboard
| Challenge: | Existing methods for relation extraction (RE) are discriminative and generative . previous studies show that discriminative models can support generative RE . |
| Approach: | They propose a framework that leverages discriminative models to produce a top-k set of candidate relations and integrates this knowledge into generative models via in-context or prompt learning. |
| Outcome: | The proposed framework achieves state-of-the-art on five widely used RE benchmarks. |
Beyond A Single AI Cluster: A Survey of Decentralized LLM Training (2025.emnlp-main)
Copied to clipboard
| Challenge: | Decentralized LLM training leverages dispersed resources at varying scales. |
| Approach: | They propose a resource-driven paradigm that leverages dispersed resources across clusters, datacenters and even regions. |
| Outcome: | The proposed model scales are 175 billion to 660 billion parameters, and the exponential growth in computational requirements poses significant challenges. |
Ensure the Correctness of the Summary: Incorporate Entailment Knowledge into Abstractive Sentence Summarization (C18-1)
Copied to clipboard
| 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. |
latent-GLAT: Glancing at Latent Variables for Parallel Text Generation (2022.acl-long)
Copied to clipboard
| Challenge: | Recent advances in text generation have limited applications due to multimodality problem. |
| Approach: | They propose a method which uses latent variables to capture word categorical information and invoke an advanced curriculum learning technique to overcome multi-modality problem. |
| Outcome: | The proposed method outperforms strong baselines without an autoregressive model, which further broadens the application scenarios of the parallel decoding paradigm. |
When is Char Better Than Subword: A Systematic Study of Segmentation Algorithms for Neural Machine Translation (2021.acl-short)
Copied to clipboard
| Challenge: | Subword segmentation algorithms can produce sub-optimal segmentation when the target language is rich in morphological changes or there is not enough data for learning compact composition rules. |
| Approach: | They compare character-based and subword-based neural machine translation systems . they find character-driven models are better at handling morphological phenomena . |
| Outcome: | The character-based models are better at handling morphological phenomena, generating rare and unknown words, and more suitable for transferring to unseen domains. |
Beyond the Individual: Virtualizing Multi-Disciplinary Reasoning for Clinical Intake via Collaborative Agents (2026.findings-acl)
Copied to clipboard
Huangwei Chen, Wu Li, Junhao Jia, Yining Chen, Xiaotao Pang, Ya-Long Chen, Li Gonghui, Haishuai Wang, Jiajun Bu, Lei Wu
| Challenge: | Initial outpatient consultations are costly and difficult to scale to real-time intake. |
| Approach: | They propose a synchronous virtual MDT framework that formalizes the consultation state using a structured SOAP representation, separating evidence collection from diagnostic reasoning to improve traceability and bias control. |
| Outcome: | The proposed framework outperforms state-of-the-art models on ClinicalBench and a real-world RAPID-IPN dataset in documentation quality and consultation capability. |
MT-PATCHER: Selective and Extendable Knowledge Distillation from Large Language Models for Machine Translation (2024.naacl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have shown their strong ability in the field of machine translation, yet they suffer from high computational cost and latency. |
| Approach: | They propose a framework which transfers knowledge from LLMs to existing MT models in a selective, comprehensive and proactive manner. |
| Outcome: | The proposed framework transfers knowledge from LLMs to existing MT models in a selective, comprehensive and proactive manner. |
Multimodal Sentence Summarization via Multimodal Selective Encoding (2020.coling-main)
Copied to clipboard
| 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 . |
Unlocking Instructive In-Context Learning with Tabular Prompting for Relational Triple Extraction (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing methods for relational triple extraction (RTE) are unnatural and recast RTE tasks to text-to-text prompting formats. |
| Approach: | They propose a tabular prompting for RTE which frames RTE task into a table generation task and propose an instructive in-context learning which only selects and annotates samples considering triple semantics in massive unlabeled samples. |
| Outcome: | The proposed prompting for RTE with TableIE achieves state-of-the-art performance compared to other methods . the proposed prompts are based on off-the shelf LLMs and are scalable to multiple scenarios . |
Towards Scalable Lightweight GUI Agents via Multi-role Orchestration (2026.findings-acl)
Copied to clipboard
Ziwei Wang, Junjie Zheng, Leyang Yang, Sheng Zhou, Xiaoxuan Tang, Fang Zhouhua, Zhiwei Liu, Dajun Chen, Yong Li, Jiajun Bu
| Challenge: | Advanced GUI agents suffer from prohibitive deployment costs on resource-constrained devices. |
| Approach: | They propose a lightweight GUI agent with GUI-specific knowledge and task scalability . LAMO-3B supports monolithic execution and MAS-style orchestration . |
| Outcome: | The proposed GUI agent LAMO-3B supports monolithic execution and MAS-style orchestration. |
From Scores to Preferences: Redefining Evaluation Paradigm for Speech Quality Reward Modeling (2026.findings-acl)
Copied to clipboard
Yifei Cao, Changhao Jiang, Jiabao Zhuang, Jiajun Sun, Ming Zhang, Zhiheng Xi, Hui Li, Shihan Dou, Yuran Wang, Yunke Zhang, Tao Ji, Tao Gui, Qi Zhang, Xuanjing Huang
| Challenge: | Experimental results show that the MOS-aware GRM significantly improves fine-grained speech quality discrimination. |
| Approach: | They propose a MOS-aware reward model that incorporates MOS gap into reward function during reinforcement learning. |
| Outcome: | The proposed model significantly improves fine-grained speech quality discrimination. |
UCoder: Unsupervised Code Generation by Internal Probing of Large Language Models (2026.findings-acl)
Copied to clipboard
Jiajun Wu, Jian Yang, Wei Zhang, Linzheng Chai, Yuchi Ma, Ensheng Shi, Yuqing Ma, Zhoujun Li, Xianglong Liu
| Challenge: | Large language models (LLMs) have demonstrated remarkable capabilities in code generation tasks, but their effectiveness relies on supervised training with extensive labeled data and computational resources. |
| Approach: | They propose an unsupervised method that leverages Internal Probing of Large language models for Code generation without any external corpus, even unlabeled code snippets. |
| Outcome: | The proposed method can achieve competitive performance compared to supervised approaches while reducing the dependency on labeled data and computational resources. |
Mitigating Object Hallucinations in MLLMs via Multi-Frequency Perturbations (2025.findings-emnlp)
Copied to clipboard
Shuo Li, Jiajun Sun, Guodong Zheng, Xiaoran Fan, Yujiong Shen, Yi Lu, Zhiheng Xi, Yuming Yang, Wenming Tan, Tao Ji, Tao Gui, Qi Zhang, Xuanjing Huang
| Challenge: | Multimodal large language models have demonstrated remarkable performance in visual-language tasks, but their authenticity is often compromised by object hallucinations. |
| Approach: | They propose a multi-frequency perturbation method that leverages both low-frequency and high-frequency features of images to perturb visual feature representations and explicitly suppress redundant frequency-domain features during inference. |
| Outcome: | The proposed method significantly mitigates object hallucinations across various model architectures. |
LAGCL4Rec: When LLMs Activate Interactions Potential in Graph Contrastive Learning for Recommendation (2025.findings-emnlp)
Copied to clipboard
Leqi Zheng, Chaokun Wang, Canzhi Chen, Jiajun Zhang, Cheng Wu, Zixin Song, Shannan Yan, Ziyang Liu, Hongwei Li
| Challenge: | Traditional contrastive learning methods treat negative feedback as equally hard or easy, ignoring informative semantic difficulty during training. |
| Approach: | They propose a framework leveraging Large Language Models to Activate interactions in Graph Contrastive Learning for Recommendation. |
| Outcome: | The proposed framework outperforms state-of-the-art benchmarks on multiple benchmarks. |
Embodied-Reasoner: Synergizing Visual Search, Reasoning, and Action for Embodied Interactive Tasks (2026.acl-long)
Copied to clipboard
Wenqi Zhang, Mengna Wang, Gangao Liu, Huixin Xu, Yiwei Jiang, Yongliang Shen, Guiyang Hou, Zhe Zheng, Hang Zhang, Xin Li, Jiajun Liu, Weiming Lu, Peng Li, Yueting Zhuang
| Challenge: | Recent advances in reasoning models have demonstrated remarkable capabilities on mathematical and coding tasks, but their effectiveness in embodied domains remains largely unexplored. |
| Approach: | They propose a reasoning model for interactive embodied tasks that synthesizes 9.3k coherent Observation-Thought-Action trajectories containing 64k ego-centric images and 90k diverse reasoning processes. |
| Outcome: | The proposed model outperforms existing visual reasoning models by +9%, 24%, and +13% on long-horizon tasks. |
CoTJudger: A Graph-Driven Framework for Automatic Evaluation of Chain-of-Thought Efficiency and Redundancy in LRMs (2026.findings-acl)
Copied to clipboard
Siyi Li, Jiajun Shi, Shiwen Ni, Ge Zhang, Shuaimin Li, Shijian Wang, Zhoufutu Wen, Yizhi LI, Hamid Alinejad-Rokny, Jiaheng Liu, Min Yang, Wenhao Huang
| Challenge: | Existing evaluations emphasize final accuracy or coarse token counts, and lack automated tools to separate essential logic from structural redundancy. |
| Approach: | They propose a graph-driven framework that quantifies reasoning efficiency by converting free-form CoTs into directed dependency graphs and extracting the Shortest Effective Path needed to reach a correct solution. |
| Outcome: | Evaluating 21 LRMs, the proposed framework quantifies reasoning efficiency by converting free-form CoTs into directed dependency graphs and extracting the Shortest Effective Path (SEP) needed to reach a correct solution. |
Golden Touchstone: A Comprehensive Bilingual Benchmark for Evaluating Financial Large Language Models (2025.findings-emnlp)
Copied to clipboard
Xiaojun Wu, Junxi Liu, Huan-Yi Su, Zhouchi Lin, Yiyan Qi, Chengjin Xu, Jiajun Su, Jiajie Zhong, Fuwei Wang, Saizhuo Wang, Fengrui Hua, Jia Li, Jian Guo
| Challenge: | Existing financial benchmarks suffer from limited language and task coverage, low-quality datasets, and inadequate adaptability for LLM evaluation. |
| Approach: | They propose a bilingual benchmark for financial LLMs that assesses models’ language understanding and generation capabilities. |
| Outcome: | The proposed bilingual benchmark assesses models’ language understanding and generation capabilities. |
Attention Calibration for Transformer in Neural Machine Translation (2021.acl-long)
Copied to clipboard
| Challenge: | Attention mechanisms have been ubiquitous in neural machine translation (NMT) however, many studies doubt whether highlyattended inputs have a large impact on the model outputs. |
| Approach: | They propose to introduce a mask perturbation model that automatically evaluates each input’s contribution to the model outputs. |
| Outcome: | The proposed model is more uniform at lower layers while more concentrated on the specific inputs at higher layers. |
MSMO: Multimodal Summarization with Multimodal Output (D18-1)
Copied to clipboard
| 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 . |