Recent advancements in Large Reasoning Models (LRMs), such as OpenAI's o1/o3 and DeepSeek-R1, have demonstrated remarkable performance in specialized reasoning tasks through human-like deliberative thinking and long chain-of-thought reasoning. However, our systematic evaluation across various model families (DeepSeek, Qwen, and LLaMA) and scales (7B to 32B) reveals that acquiring these deliberative reasoning capabilities significantly reduces the foundational capabilities of LRMs, including notable declines in helpfulness and harmlessness, alongside substantially increased inference costs. Importantly, we demonstrate that adaptive reasoning---employing modes like Zero-Thinking, Less-Thinking, and Summary-Thinking---can effectively alleviate these drawbacks. Our empirical insights underline the critical need for developing more versatile LRMs capable of dynamically allocating inference-time compute according to specific task characteristics.

Jiahe Guo, Xiangran Guo, Yulin Hu, Zimo Long, Xingyu Sui, Xuda Zhi, Yongbo Huang, Hao He, Weixiang Zhao, Yanyan Zhao, Bing Qin. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Emotional Support Conversation requires not only affective expression but also grounded instrumental support to provide trustworthy guidance.However, existing ESC systems and benchmarks largely focus on affective support in text-only settings, overlooking how external tools can enable factual grounding and reduce hallucination in multi-turn emotional support.We introduce TEA-Bench, the first interactive benchmark for evaluating tool-augmented agents in ESC, featuring realistic emotional scenarios, an MCP-style tool environment, and process-level metrics that jointly assess the quality and factual grounding of emotional support.Experiments on nine LLMs show that tool augmentation generally improves emotional support quality and reduces hallucination, but the gains are strongly capacity-dependent: stronger models use tools more selectively and effectively, while weaker models benefit only marginally.We further release TEA-Dialog, a dataset of toolenhanced ESC dialogues, and find that supervised fine-tuning improves in-distribution support but generalizes poorly.Our results underscore the importance of tool use in building reliable emotional support agents. 1

Weixiang Zhao, Yulin Hu, Yang Deng, Jiahe Guo, Xingyu Sui, Xinyang Han, An Zhang, Yanyan Zhao, Bing Qin, Tat-Seng Chua, Ting Liu. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Weixiang Zhao, Yulin Hu, Yang Deng, Tongtong Wu, Wenxuan Zhang, Jiahe Guo, An Zhang, Yanyan Zhao, Bing Qin, Tat-Seng Chua, Ting Liu. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Weixiang Zhao, Jiahe Guo, Yulin Hu, Yang Deng, An Zhang, Xingyu Sui, Xinyang Han, Yanyan Zhao, Bing Qin, Tat-Seng Chua, Ting Liu. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025.

Multilingual reasoning remains a significant challenge for large language models (LLMs), with performance disproportionately favoring high-resource languages. Drawing inspiration from cognitive neuroscience, which suggests that human reasoning functions largely independently of language processing, we hypothesize that LLMs similarly encode reasoning and language as separable components that can be disentangled to enhance multilingual reasoning. To evaluate this, we perform a causal intervention by ablating language-specific representations at inference time. Experiments on 10 open-weight LLMs spanning 11 typologically diverse languages show that this language-specific ablation consistently boosts multilingual reasoning performance. Layer-wise analyses further confirm that language and reasoning representations can be effectively disentangled throughout the model, yielding improved multilingual reasoning capabilities, while preserving top-layer language features remains essential for maintaining linguistic fidelity. Compared to post-training methods such as supervised fine-tuning or reinforcement learning, our training-free language-reasoning disentanglement achieves comparable or superior results with minimal computational overhead. These findings shed light on the internal mechanisms underlying multilingual reasoning in LLMs and suggest a lightweight and interpretable strategy for improving cross-lingual generalization.

Personalized alignment is essential for enabling large language models (LLMs) to engage effectively in user-centric dialogue. While recent prompt-based and offline optimization methods offer preliminary solutions, they fall short in cold-start scenarios and long-term personalization due to their inherently static and shallow designs. In this work, we introduce the Reinforcement Learning for Personalized Alignment (RLPA) framework, in which an LLM interacts with a simulated user model to iteratively infer and refine user profiles through dialogue. The training process is guided by a dual-level reward structure: the Profile Reward encourages accurate construction of user representations, while the Response Reward incentivizes generation of responses consistent with the inferred profile. We instantiate RLPA by fine-tuning Qwen-2.5-3B-Instruct, resulting in Qwen-RLPA, which achieves state-of-the-art performance in personalized dialogue. Empirical evaluations demonstrate that Qwen-RLPA consistently outperforms prompting and offline fine-tuning baselines, and even surpasses advanced commercial models such as Claude-3.5 and GPT-4o. Further analysis highlights Qwen-RLPA's robustness in reconciling conflicting user preferences, sustaining long-term personalization and delivering more efficient inference compared to recent reasoning-focused LLMs. These results emphasize the potential of dynamic profile inference as a more effective paradigm for building personalized dialogue systems.

This paper investigates a movable antenna (MA)-aided integrated sensing and communication (ISAC) system using non-orthogonal multiple access (NOMA). A base station (BS) configured with a two-dimensional MA array simultaneously serves multiple communication users and sensing multiple targets. To enhance system capacity, superimposed symbols are transmitted to communication users, with successive interference cancellation (SIC) employed for signal decoding. Our objective is to maximize the total illumination power at the targets while satisfying the minimum signal-to-interference-plus-noise ratio (SINR) requirements for communication users. To achieve this goal, we propose an alternating optimization (AO)-based algorithm that jointly optimizes the transmit power allocation, beamforming, sensing covariance matrix, and antenna positions. Numerical results show that the MA system achieves significant improvement in illumination power compared to fixed-position antennas (FPAs), with particularly significant gains under high SINR requirements.

As wireless communication advances toward the 6G era, the demand for ultra-reliable, high-speed, and ubiquitous connectivity is driving the exploration of new degrees-of-freedom (DoFs) in communication systems. Among the key enabling technologies, Movable Antennas (MAs) integrated into Flexible Cylindrical Arrays (FCLA) have shown great potential in optimizing wireless communication by providing spatial flexibility. This paper proposes an innovative optimization framework that leverages the dynamic mobility of FCLAs to improve communication rates and overall system performance. By employing Fractional Programming (FP) for alternating optimization of beamforming and antenna positions, the system enhances throughput and resource utilization. Additionally, a novel Constrained Grid Search-Based Adaptive Moment Estimation Algorithm (CGS-Adam) is introduced to optimize antenna positions while adhering to antenna spacing constraints. Extensive simulations validate that the proposed system, utilizing movable antennas, significantly outperforms traditional fixed antenna optimization, achieving up to a 31% performance gain in general scenarios. The integration of FCLAs in wireless networks represents a promising solution for future 6G systems, offering improved coverage, energy efficiency, and flexibility.