Papers by Junyi Zhang
RubricBench: Aligning Model-Generated Rubrics with Human Standards (2026.acl-long)
Copied to clipboard
Junyi Zhou, Qiyuan Zhang, Yufei Wang, Fuyuan Lyu, Yidong Ming, Can Xu, Qingfeng Sun, Kai Zheng, Peng Kang, Xue Liu, Chen Ma
| Challenge: | Existing benchmarks lack discriminative complexity and ground-truth rubric annotations required for rigorous evaluation. |
| Approach: | They propose a curated benchmark with 1,147 pairwise comparisons to assess the reliability of rubric-based evaluation. |
| Outcome: | The proposed benchmarks show that they support diverse domains, exhibit discriminative ability, provide high-quality annotations, and include human-authored rubrics. |
RAG-Star: Enhancing Deliberative Reasoning with Retrieval Augmented Verification and Refinement (2025.naacl-long)
Copied to clipboard
| Challenge: | Existing large language models (LLMs) show exceptional problem-solving capabilities but struggle with complex reasoning tasks. |
| Approach: | They propose a novel RAG approach that integrates retrieved information to guide tree-based reasoning process based on LLMs. |
| Outcome: | The proposed approach outperforms existing methods in large language models . iteratively plans intermediate sub-queries and answers based on the LLM itself . |
ManuSearch: Democratizing Deep Search in Large Language Models with a Transparent and Open Multi-Agent Framework (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing systems with opaque architectures are limiting deep search capabilities for web-augmented large language models. |
| Approach: | They propose a transparent and modular multi-agent framework to democratize deep search for LLMs. |
| Outcome: | The proposed framework outperforms open-source systems in deep reasoning tasks. |
PERM: Psychology-grounded Empathetic Reward Modeling for Large Language Models (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing reward models evaluate empathy from a single perspective, overlooking bidirectional interaction nature of empathy. |
| Approach: | They propose a reward model that evaluates empathy from a single perspective . they propose PERM to integrate a bystander perspective to monitor overall interaction quality . |
| Outcome: | a new reward model outperforms state-of-the-art models on an emotional intelligence benchmark and an industrial daily conversation dataset. |
Federated LoRA Fine-Tuning with Pipelined Error-Mitigated Aggregation and Matrix-Wise Freezing (2026.findings-acl)
Copied to clipboard
| 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. |
LLM-Driven Completeness and Consistency Evaluation for Cultural Heritage Data Augmentation in Cross-Modal Retrieval (2025.emnlp-main)
Copied to clipboard
| Challenge: | Cross-modal retrieval is essential for interpreting cultural heritage data, but its effectiveness is limited by incomplete or inconsistent textual descriptions. |
| Approach: | They propose a data augmentation framework that enhances cross-modal retrieval performance by improving the completeness and consistency of LLM-generated descriptions. |
| Outcome: | The proposed framework improves cross-modal retrieval performance by improving completeness and consistency of LLM-generated descriptions. |
XtraGPT: Context-Aware and Controllable Academic Paper Revision via Human-AI Collaboration (2026.acl-long)
Copied to clipboard
Nuo Chen, Andre Lin HuiKai, Jiaying Wu, Junyi Hou, Zining Zhang, Qian Wang, Xidong Wang, Bingsheng He
| Challenge: | Existing systems are designed for general-purpose scientific text generation and fail to support high-quality scientific writing beyond surface-level polishing. |
| Approach: | They propose a human-AI collaboration framework for academic paper revision based on criteria-guided intent alignment and context-aware modeling. |
| Outcome: | The proposed framework outperforms existing LLMs and rivals the quality of proprietary ones. |
CLUE: A Chinese Language Understanding Evaluation Benchmark (2020.coling-main)
Copied to clipboard
Liang Xu, Hai Hu, Xuanwei Zhang, Lu Li, Chenjie Cao, Yudong Li, Yechen Xu, Kai Sun, Dian Yu, Cong Yu, Yin Tian, Qianqian Dong, Weitang Liu, Bo Shi, Yiming Cui, Junyi Li, Jun Zeng, Rongzhao Wang, Weijian Xie, Yanting Li, Yina Patterson, Zuoyu Tian, Yiwen Zhang, He Zhou, Shaoweihua Liu, Zhe Zhao, Qipeng Zhao, Cong Yue, Xinrui Zhang, Zhengliang Yang, Kyle Richardson, Zhenzhong Lan
| Challenge: | Existing language evaluation benchmarks for English are limited to English . lack of such benchmarks makes it difficult to replicate success in other languages . |
| Approach: | They introduce a large-scale Chinese language understanding evaluation benchmark . the benchmark uses a set of current state-of-the-art pre-trained Chinese models . |
| Outcome: | The first large-scale Chinese Language Understanding Evaluation (CLUE) benchmark is released . the benchmark evaluates models across a wide range of tasks on original Chinese text . existing language evaluation benchmarks are mostly limited to English . |
Empathy in Diversity: Personalized Depression and Anxiety Therapy via Dialogue State Tracking and Patient-Aware Planning (2026.acl-long)
Copied to clipboard
Xinwei Yang, Junyi Fan, Yuqing Liu, Jiaxuan Wang, Jiashuai Zhang, Hongru Liang, Wenqiang Lei, Yao Song
| Challenge: | Recent efforts have turned to large language models (LLMs) as therapeutic agents for psychological therapy tasks, yet robustness across diverse patients remains underexplored. |
| Approach: | They propose a realistic role-play protocol for evaluating therapeutic dialogue agents and a de-identified, expert-annotated corpus of therapist–patient dialogues. |
| Outcome: | The proposed framework outperforms baselines on therapeutic outcomes and dialogue quality while improving conversational efficiency. |
MemSearch-o1: Empowering Large Language Models with Reasoning-Aligned Memory Growth in Agentic Search (2026.acl-long)
Copied to clipboard
Sheng Zhang, Junyi Li, Yingyi Zhang, Pengyue Jia, Yichao Wang, Xiaowei Qian, Wenlin Zhang, Maolin Wang, Yong Liu, Xiangyu Zhao
| Challenge: | Existing methods for memory management struggle to capture fine-grained semantic relations between queries and documents. |
| Approach: | They propose a framework for reasoning and agentic search that grows fine-grained memory fragments from seed tokens from queries, then retraces and deep refines the memory via a contribution function. |
| Outcome: | Experiments on eight benchmark datasets show that MemSearch-o1 significantly mitigates memory dilution and more effectively activates reasoning potential of diverse LLMs. |
BRIEF-Pro: Universal Context Compression with Short-to-Long Synthesis for Fast and Accurate Multi-Hop Reasoning (2026.findings-acl)
Copied to clipboard
| Challenge: | Experiments show that BRIEF-Pro generates more concise and relevant summaries, enhancing performance across small, large, and proprietary language models. |
| Approach: | They propose a universal, lightweight compressor that distills relevant evidence from retrieved documents into a concise summary for seamless integration into in-context RAG. |
| Outcome: | Experiments on four open-domain multi-hop question-answering datasets show that BRIEF-Pro generates more concise and relevant summaries, enhancing performance across small, large, and proprietary language models. |
SpeechUT: Bridging Speech and Text with Hidden-Unit for Encoder-Decoder Based Speech-Text Pre-training (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing models for pre-training text and speech are based on unlabeled audio data. |
| Approach: | They propose a unified-modal speech-unit-text pre-training model that connects speech encoders and text decoders with a shared unit encoder. |
| Outcome: | The proposed model improves on automatic speech recognition and speech translation tasks and achieves state-of-the-art performance on both the LibriSpeech ASR and MuST-C ST tasks. |
Confidence-Calibrated Small-Large Language Model Collaboration for Cost-Efficient Reasoning (2026.eacl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) have superior reasoning capabilities compared to small language models, but incur substantially higher inference costs. |
| Approach: | They propose a system that cascades an LLM with an SLM to achieve a balance between accuracy and cost in complex reasoning tasks. |
| Outcome: | The proposed system improves the SLM’s reasoning ability and confidence calibration across diverse datasets and model backbones. |
You Only Query Twice: Multimodal Rumor Detection via Evidential Evaluation from Dual Perspectives (2025.coling-main)
Copied to clipboard
| Challenge: | Existing rumor detectors exhibit limitations in fully exploiting responses to the source tweet as essential public opinions, and in explaining and indicating the reliability of the results obtained. Existing research mainly combats this with content and response-based detection methods. |
| Approach: | They propose a Large Language Model with both multimodal source content and the corresponding response set to extract contrasting evidence to enable maximal utilization of informative responses. |
| Outcome: | The proposed approach can indicate the model’s uncertainty (i.e., reliability) of the results. |
Adversarial Preference Learning for Robust LLM Alignment (2025.findings-acl)
Copied to clipboard
Yuanfu Wang, Pengyu Wang, Chenyang Xi, Bo Tang, Junyi Zhu, Wenqiang Wei, Chen Chen, Chao Yang, Jingfeng Zhang, Chaochao Lu, Yijun Niu, Keming Mao, Zhiyu Li, Feiyu Xiong, Jie Hu, Mingchuan Yang
| Challenge: | Modern language models rely on Reinforcement Learning from Human Feedback (RLHF) to encourage safe behaviors, but they remain vulnerable to adversarial attacks due to three key limitations: (1) the inefficiency and high cost of human annotation; (2) the vast diversity of potential adversarials; and (3) the risk of feedback bias and reward hacking. |
| Approach: | They propose an iterative adversarial training method that incorporates three key innovations to address these challenges. |
| Outcome: | Experiments on Mistral-7B-Instruct-v0.3 show that the proposed method significantly enhances robustness and reduces harmful outputs from 5.88% to 0.43%. |
Dunhuang-Bench: How Well Do MLLMs Understand Cultural Heritage? (2026.findings-acl)
Copied to clipboard
| Challenge: | Recent advances in Multimodal Large Language Models (MLLMs) have led to extensive evaluations on Chinese cultural benchmarks. |
| Approach: | They construct a large-scale benchmark comprising 486 images and 22,970 QA pairs to evaluate MLLMs' cultural understanding. |
| Outcome: | The proposed benchmark incorporates three task formats to evaluate MLLMs’ cultural understanding: Question Answering with Text Description, Multi-turn Dialogue, and Question Answers with Choices. |
SpeechT5: Unified-Modal Encoder-Decoder Pre-Training for Spoken Language Processing (2022.acl-long)
Copied to clipboard
Junyi Ao, Rui Wang, Long Zhou, Chengyi Wang, Shuo Ren, Yu Wu, Shujie Liu, Tom Ko, Qing Li, Yu Zhang, Zhihua Wei, Yao Qian, Jinyu Li, Furu Wei
| Challenge: | Existing work shows that pre-trained models can improve in various natural language processing tasks. |
| Approach: | They propose a unified-modal encoder-decoder framework that pre-trains speech-text representations using large-scale unlabeled speech and text data. |
| Outcome: | The proposed framework is superior to existing models on speech-to-text processing tasks. |
Equal Truth: Rumor Detection with Invariant Group Fairness (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing rumor detection methods rarely consider fairness issues inherent in the model . this can lead to biased predictions across stakeholder groups, undermining their detection effectiveness . |
| Approach: | They propose a framework to address fairness issues inherent in rumor detection models . they perform unsupervised partitioning to dynamically identify potential unfair data patterns . then, they apply invariant learning to these partitions to extract fair and informative feature representations . |
| Outcome: | The proposed method outperforms strong baselines regarding detection and fairness performance . it also shows robust performance on out-of-distribution samples . |
JailMeter: An Evidence-Based Evaluation Framework for Jailbreak Attacks on Large Language Models (2026.findings-acl)
Copied to clipboard
Qingjia Huang, Jingyu Zhang, Jianguo Wu, Yakai Li, Weijuan Zhang, Yankai Rong, Junyi Yao, Shengzhi Zhang, Xiaoqi Jia
| Challenge: | Currently, evaluation criteria and methods used for jailbreak effectiveness are inconsistent. |
| Approach: | They propose a framework to measure jailbreak effectiveness using a model that filters out jailbreak noise while preserving the original malicious question. |
| Outcome: | The proposed framework outperforms existing evaluation methods on a challenging benchmark containing 330 human-labeled, non-rejected jailbreak instances. |
Small Agent Can Also Rock! Empowering Small Language Models as Hallucination Detector (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing studies on hallucination detection rely heavily on closed-source LLMs such as GPT-4. |
| Approach: | They propose an LLM-based agent framework called HaluAgent that integrates LLMs, multi-functional toolbox and a memory mechanism for hallucination detection. |
| Outcome: | The proposed framework integrates the LLM, multi-functional toolbox, and can detect hallucinations on Chinese and English datasets. |