Author name cluster

Xiaoming Fu

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

6 papers

1 author row

AAAI Conference 2026 Conference Paper

ESCA: An Emotional Support Conversation Agent for Enhancing Reasonable Strategy Planning and Effective Expression

Jing Li
Yanxin Luo
Donghong Han
Yimeng Zhan
Xiaoming Fu
Baiyou Qiao
Gang Wu

Emotional Support Conversation (ESC) aims to alleviate individuals’ negative emotions through multi-turn dialogues, where effective strategy planning and response generation are essential. However, existing methods often suffer from limitations in both planning reasonable support strategies and effectively expressing them in responses. To the end, we propose a novel LLM-based Emotional Support Conversation Agent (ESCA) with a plug-in strategy planner and a strategy-aligned prompt generator. The strategy planner cooperates with four aspects of the seeker’s state, including emotion intensity, trust degree, dialogue behavior, and stage of change, to enhance the rationality and effectiveness of the strategy prediction. To ensure that predicted strategies are better conveyed, the prompt generator integrates strategy-aligned instructions, knowledge, and context to generate the soft prompt for guiding the LLM to generate supportive responses. In addition to supervised fine-tuning, the prompt generator is further optimized by reinforcement learning. Experimental results demonstrate that ESCA significantly improves both response quality and the success rate of achieving the ESC task goal.

PDF Details DOI

IJCAI Conference 2025 Conference Paper

COGRASP: Co-Occurrence Graph Based Stock Price Forecasting

Zhengze Li
Zilin Song
Tingting Yuan
Xiaoming Fu

Forecasting stock prices is complex and challenging. Uncovering correlations among stocks has proven to enhance stock price forecasting. However, existing correlation discovery methods, such as concept-based methods, are slow, inaccurate, and limited by their reliance on predefined concepts and manual analysis. In this paper, we propose COGRASP, a novel approach for stock price forecasting that constructs stock co-occurrence graphs automatically by analyzing rapidly updated sources such as reports, newspapers, and social media. Besides, we aggregate forecasts across multiple timescales (i. e. , long-, medium-, and short-term) to capture multi-timescale trends fluctuations, thereby enhancing price forecasting accuracy. In experiments with real-world open-source stock market data, COGRASP outperforms state-of-the-art methods.

PDF Details DOI

AAMAS Conference 2025 Conference Paper

ReSCOM: Reward-Shaped Curriculum for Efficient Multi-Agent Communication Learning

Xinghai Wei
Tingting Yuan
Jie Yuan
Dongxiao Liu
Xiaoming Fu

Communication enhances collaboration among artificial intelligence agents. Given the conflicts between limited communication resources and communication needs, learning effective communication strategies is essential. We observe that incorporating learning to communicate can complicate mastering primary tasks. This is due to the uncertainty in information acquisition during the learning process, which can lead to an unstable environment for primary tasks. In this paper, we introduce ReSCOM, an efficient joint learning framework that combines learning-to-communicate with primary tasks. ReSCOM progressively adjusts the learning emphasis through reward-shaped curriculums, allowing agents to shift their focus from primary tasks and basic communication tasks (e. g. , how to encode) to advanced communication strategies (e. g. , determining when it is worthwhile to communicate). This approach minimizes the impact on the learning efficiency of primary tasks while simultaneously facilitating communication learning. We evaluate ReSCOM against state-of-the-art methods across various tasks, and experimental results demonstrate its effectiveness.

PDF

NeurIPS Conference 2025 Conference Paper

SmallKV: Small Model Assisted Compensation of KV Cache Compression for Efficient LLM Inference

Yi Zhao
Yajuan Peng
Nguyen Cam-Tu
Zuchao Li
Xiaoliang Wang
Hai Zhao
Xiaoming Fu

KV cache eviction has emerged as an effective solution to alleviate resource constraints faced by LLMs in long-context scenarios. However, existing token-level eviction methods often overlook two critical aspects: (1) their irreversible eviction strategy fails to adapt to dynamic attention patterns during decoding (the saliency shift problem), and (2) they treat both marginally important tokens and truly unimportant tokens uniformly, despite the collective significance of marginal tokens to model performance (the marginal information over-compression problem). To address these issues, we design two compensation mechanisms based on the high similarity of attention matrices between LLMs with different scales. We propose SmallKV, a small model assisted compensation method for KV cache compression. SmallKV can maintain attention matching between different-scale LLMs to: 1) assist the larger model in perceiving globally important information of attention; and 2) use the smaller model’s attention scores to approximate those of marginal tokens in the larger model. Extensive experiments on benchmarks including GSM8K, BBH, MT-Bench, and LongBench demonstrate the effectiveness of SmallKV. Moreover, efficiency evaluations show that SmallKV achieves 1. 75 - 2. 56 times higher throughput than baseline methods, highlighting its potential for efficient and performant LLM inference in resource constrained environments.

PDF Details

AAAI Conference 2023 Conference Paper

DACOM: Learning Delay-Aware Communication for Multi-Agent Reinforcement Learning

Tingting Yuan
Hwei-Ming Chung
Jie Yuan
Xiaoming Fu

Communication is supposed to improve multi-agent collaboration and overall performance in cooperative Multi-agent reinforcement learning (MARL). However, such improvements are prevalently limited in practice since most existing communication schemes ignore communication overheads (e.g., communication delays). In this paper, we demonstrate that ignoring communication delays has detrimental effects on collaborations, especially in delay-sensitive tasks such as autonomous driving. To mitigate this impact, we design a delay-aware multi-agent communication model (DACOM) to adapt communication to delays. Specifically, DACOM introduces a component, TimeNet, that is responsible for adjusting the waiting time of an agent to receive messages from other agents such that the uncertainty associated with delay can be addressed. Our experiments reveal that DACOM has a non-negligible performance improvement over other mechanisms by making a better trade-off between the benefits of communication and the costs of waiting for messages.

PDF Details DOI

IJCAI Conference 2021 Conference Paper

Temporal Heterogeneous Information Network Embedding

Hong Huang
Ruize Shi
Wei Zhou
Xiao Wang
Hai Jin
Xiaoming Fu

Heterogeneous information network (HIN) embedding, learning the low-dimensional representation of multi-type nodes, has been applied widely and achieved excellent performance. However, most of the previous works focus more on static heterogeneous networks or learning node embedding within specific snapshots, and seldom attention has been paid to the whole evolution process and capturing all temporal dynamics. In order to fill the gap of obtaining multi-type node embeddings by considering all temporal dynamics during the evolution, we propose a novel temporal HIN embedding method (THINE). THINE not only uses attention mechanism and meta-path to preserve structures and semantics in HIN but also combines the Hawkes process to simulate the evolution of the temporal network. Our extensive evaluations with various real-world temporal HINs demonstrate that THINE achieves state-of-the-art performance in both static and dynamic tasks, including node classification, link prediction, and temporal link recommendation.

PDF Details DOI