Papers by Xinyu Tang

Memory-Efficient Backpropagation for Fine-Tuning LLMs on Resource-Constrained Mobile Devices (2025.emnlp-industry)

Copied to clipboard

Challenge:	Existing work on memory-efficient on-device fine-tuning of large language models with backpropagation has focused on approximating gradients with zeroth-order optimization (ZO).
Approach:	They propose a memory-efficient implementation of backpropagation on mobile devices that allows flexible trade-offs between memory usage and compute time while converging faster.
Outcome:	The proposed method can fine-tune LLMs with backpropagation using less than 1GB of memory while achieving better performance than the baseline.

Emergent Relational Order in LLM Agent Societies: From Collective Affect to Authority Stratification (2026.findings-acl)

Copied to clipboard

Zhiyuan Ji, Xinyu Chen, Ziqi Dai, Shiyun Tang, Chunyu Wei, Yueguo Chen

Linguistic Theories

Challenge:	Fei Xiaotong’s Differential Order Pattern characterizes rural society as egocentric and relationally graded, with cooperation attenuating over social distance.
Approach:	They propose a multi-agent framework grounded in Affect Control Theory, Social Identity Theory, and Durkheimian collective affect.
Outcome:	Extensive simulations support interpreting Differential Order as a structure-sensitive emergent outcome of general social mechanisms.

Rethinking Sample Polarity in Reinforcement Learning with Verifiable Rewards (2026.acl-long)

Copied to clipboard

Xinyu Tang, Yuliang Zhan, Zhixun Li, Xin Zhao, Zhenduo Zhang, Zujie Wen, Zhiqiang Zhang, Jun Zhou

Reasoning

Challenge:	Large reasoning models are typically trained using reinforcement learning with verifiable reward (RLVR) positive and negative self-generated rollouts are used to update the model's policy . positive samples sharpen existing correct reasoning patterns, while negative samples encourage exploration of new reasoning paths.
Approach:	They propose a method that allocates advantage signals to key tokens across different polarities.
Outcome:	The proposed method improves the ability of large reasoning models to learn from their own generated rollouts.

Debias NLU Datasets via Training-free Perturbations (2023.findings-emnlp)

Copied to clipboard

Qi Guo, Yuanhang Tang, Yawen Ouyang, Zhen Wu, Xinyu Dai

Robustness in NLP

Challenge:	Existing approaches to debiase NLU models capture biased features that are independent of the task but spuriously correlated to labels.
Approach:	They propose a framework that conducts training-free perturbations on samples containing biased features to Debias NLU Datasets.
Outcome:	The proposed framework shows competitive performance with previous state-of-the-art debiasing strategies.

DAWN-ICL: Strategic Planning of Problem-solving Trajectories for Zero-Shot In-Context Learning (2025.naacl-long)

Copied to clipboard

Xinyu Tang, Xiaolei Wang, Xin Zhao, Ji-Rong Wen

Responsible & Trustworthy NLP

Challenge:	Existing methods to conduct in-context learning without using human-annotated demonstrations are unreliable and lead to error accumulation.
Approach:	They propose a method to conduct in-context learning without using human-annotated demonstrations.
Outcome:	The proposed method outperforms existing methods using human-annotated demonstrations.

CultureSynth: A Hierarchical Taxonomy-Guided and Retrieval-Augmented Framework for Cultural Question-Answer Synthesis (2025.findings-emnlp)

Copied to clipboard

Xinyu Zhang, Pei Zhang, Shuang Luo, Jialong Tang, Yu Wan, Baosong Yang, Fei Huang

Information Retrieval

Challenge:	Cultural competence is defined as the ability to understand and adapt to multicultural contexts.
Approach:	They propose a framework that uses a hierarchical multilingual taxonomy and a Retrieval-Augmented Generation to synthesize culturally relevant question-answer pairs.
Outcome:	The proposed framework contains a hierarchical multilingual taxonomy covering 12 primary and 130 secondary topics and a Retrieval-Augmented Generation (RAG)-based methodology leveraging factual knowledge to synthesize culturally relevant question-answer pairs.

Glyph: Scaling Context Windows via Visual-Text Compression (2026.acl-long)

Copied to clipboard

Jiale Cheng, Yusen Liu, Xinyu Zhang, Yulin Fei, Wenyi Hong, Ruiliang Lyu, Weihan Wang, Zhe Su, Xiaotao Gu, Xiao Liu, Yushi Bai, Jie Tang, Hongning Wang, Minlie Huang

Multimodality

Challenge:	Large language models (LLMs) traditionally represent text as sequences of discrete tokens . a long-context scaling problem requires processing more tokens more efficiently .
Approach:	They propose a framework that renders long texts into compact visual pages and processes them with a vision-language model.
Outcome:	The proposed framework renders long texts into compact visual pages and processes them with a vision-language model.

IIET: Efficient Numerical Transformer via Implicit Iterative Euler Method (2025.emnlp-main)

Copied to clipboard

Xinyu Liu, Bei Li, Jiahao Liu, Junhao Ruan, Kechen Jiao, Hongyin Tang, Jingang Wang, Tong Xiao, JingBo Zhu

Green & Sustainable NLP

Challenge:	High-order numerical methods enhance performance in tasks like NLP but introduce a performance-efficiency trade-off due to increased computational overhead.
Approach:	They propose an iterative implicit Euler Transformer which simplifies high-order numerical methods by iterating implicit Eule.
Outcome:	The proposed method improves accuracy and reduces inference overhead by 55% while retaining 99.4% of the original task accuracy.

Showing Your Work Doesn’t Always Work (2020.acl-main)

Copied to clipboard

Raphael Tang, Jaejun Lee, Ji Xin, Xinyu Liu, Yaoliang Yu, Jimmy Lin

Challenge:	countless experimental papers lack empirical rigor, disregarding necessities such as statistical significance tests and computational environments.
Approach:	They propose to report the expected validation effectiveness of the best-tuned model with respect to the computational budget.
Outcome:	The proposed model favors negative errors and yields poor bootstrapped confidence intervals, the authors argue . they find that the proposed model is biased and uses error-prone assumptions .

RAGLAB: A Modular and Research-Oriented Unified Framework for Retrieval-Augmented Generation (2024.emnlp-demo)

Copied to clipboard

Xuanwang Zhang, Yun-Ze Song, Yidong Wang, Shuyun Tang, Xinfeng Li, Zhengran Zeng, Zhen Wu, Wei Ye, Wenyuan Xu, Yue Zhang, Xinyu Dai, Shikun Zhang, Qingsong Wen

Information Retrieval

Challenge:	Existing research on Retrieval Augmented Generation (RAG) does not address the problem of hallucinations and real-time updating of knowledge.
Approach:	They propose a modular open-source library to equip LLMs with external knowledge.
Outcome:	The proposed approach reduces the need for expensive open-source tools and lacks fair comparisons between novel RAG algorithms.

Rethinking the Evaluation for Conversational Recommendation in the Era of Large Language Models (2023.emnlp-main)

Copied to clipboard

Xiaolei Wang, Xinyu Tang, Xin Zhao, Jingyuan Wang, Ji-Rong Wen

Dialogue Systems & Conversational Agents

Challenge:	Existing evaluation protocols for large language models (LLMs) are inadequate for conversational recommender systems.
Approach:	They propose an evaluation approach based on LLMs that harnesses LLM-based user simulators to evaluate ChatGPT's performance.
Outcome:	The proposed evaluation approach can simulate various system-user interaction scenarios.

Unlocking General Long Chain-of-Thought Reasoning Capabilities of Large Language Models via Representation Engineering (2025.acl-long)

Copied to clipboard

Xinyu Tang, Xiaolei Wang, Zhihao Lv, Yingqian Min, Xin Zhao, Binbin Hu, Ziqi Liu, Zhiqiang Zhang

Semantic Text Processing

Challenge:	Existing work finds that long CoT reasoning can be efficiently elicited by tuning on only a few examples and can easily transfer to other tasks.
Approach:	They propose a representation engineering method to unleash the general long CoT reasoning capabilities of LLMs.
Outcome:	The proposed method is effective in in-domain and cross-domain scenarios.

Improving Context Fidelity via Native Retrieval-Augmented Reasoning (2025.emnlp-main)

Copied to clipboard

Suyuchen Wang, Jinlin Wang, Xinyu Wang, Shiqi Li, Xiangru Tang, Sirui Hong, Xiao-Wen Chang, Chenglin Wu, Bang Liu

Reasoning

Challenge:	Existing approaches to fidelity to contexts rely on expensive supervised fine-tuning to generate evidence post-answer or train models to perform web searches without improving utilization of the given context.
Approach:	They propose a native retrieval-augmented reasoning framework that integrates in-context evidence with the model’s own retrieval capabilities.
Outcome:	The proposed approach outperforms supervised fine-tuning, retrieval-augmented generation methods, and external retrieval solutions on multiple real-world and counterfactual QA benchmarks.

NavA3: Understanding Any Instruction, Navigating Anywhere, Finding Anything (2026.acl-long)

Copied to clipboard

Lingfeng Zhang, Xiaoshuai Hao, Yingbo Tang, Haoxiang Fu, Xinyu Zheng, Pengwei Wang, Zhongyuan Wang, Wenbo Ding, Shanghang Zhang

Reasoning

Challenge:	Existing embodied navigation methods struggle with such tasks due to their limitations in comprehending high-level human instructions and localizing objects with an open vocabulary.
Approach:	They propose a hierarchical framework for long-horizon navigation that integrates human instructions with 3D scene views.
Outcome:	The proposed model achieves SOTA results and can complete long-horizon navigation tasks across different robot embodiments in real-world environments.