Author name cluster

Weixiang Zhao

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.

10 papers

1 author row

AAAI Conference 2026 Conference Paper

CultureRL: Internalizing Cultural Principles in Large Language Models via Norm-Driven Reinforcement Learning

Weixiang Zhao
Haozhen Li
Yanyan Zhao
Haixiao Liu
Biye Li
Ting Liu
Bing Qin

As large language models (LLMs) are increasingly deployed across culturally diverse regions, ensuring that their responses align with users’ cultural norms has become a critical challenge. Existing approaches to cultural alignment primarily rely on prompting or data-augmentation-based supervised finetuning, which teach models to follow norms indirectly through example-based supervision. However, these methods are difficult to scale and often fail to generalize, particularly in low-resource cultural settings. In this work, we propose CultureRL, a culture-norm-driven reinforcement learning framework that directly encodes cultural principles into model behavior. Rather than relying on output imitation, CultureRL provides normative feedback during training, enabling the model to internalize high-level cultural rules. It consists of two key components: (1) Norm Pool Construction (NPC), which clusters data from the World Values Survey into abstract cultural concepts to form a structured and retrievable norm pool; and (2) Norm Cluster-based Reward Mechanism (NCRM), which retrieves the relevant norm for each input and uses an external reward model to assess conformity, guiding model updates through cultural alignment. We evaluate CultureRL in both one-for-one (per-culture) and one-for-all (multi-culture) settings across nine cultures and three benchmarks. Results show that CultureRL consistently outperforms strong baselines, especially in terms of cultural consistency and adaptability.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Trade-offs in Large Reasoning Models: An Empirical Analysis of Deliberative and Adaptive Reasoning over Foundational Capabilities

Weixiang Zhao
Xingyu Sui
Jiahe Guo
Yulin Hu
Yang Deng
Yanyan Zhao
Xuda Zhi
Yongbo Huang

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.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

L-MTP: Leap Multi-Token Prediction Beyond Adjacent Context for Large Language Models

Xiaohao Liu
Xiaobo Xia
Weixiang Zhao
Manyi Zhang
Xianzhi Yu
Xiu Su
Shuo Yang
See-Kiong Ng

Large language models (LLMs) have achieved notable progress. Despite their success, next-token prediction (NTP), the dominant method for LLM training and inference, is constrained in both contextual coverage and inference efficiency due to its inherently sequential process. To overcome these challenges, we propose leap multi-token prediction~(L-MTP), an innovative token prediction method that extends the capabilities of multi-token prediction (MTP) by introducing a leap-based mechanism. Unlike conventional MTP, which generates multiple tokens at adjacent positions, L-MTP strategically skips over intermediate tokens, predicting non-sequential ones in a single forward pass. This structured leap not only enhances the model's ability to capture long-range dependencies but also enables a decoding strategy specially optimized for non-sequential leap token generation, effectively accelerating inference. We theoretically demonstrate the benefit of L-MTP in improving inference efficiency. Experiments across diverse benchmarks validate its merit in boosting both LLM performance and inference speed. The source code is available at https: //github. com/Xiaohao-Liu/L-MTP.

PDF Details

NeurIPS Conference 2025 Conference Paper

On Reasoning Strength Planning in Large Reasoning Models

Leheng Sheng
An Zhang
Zijian Wu
Weixiang Zhao
Changshuo Shen
Zhang Yi
Xiang Wang
Tat-Seng Chua

Recent studies empirically reveal that large reasoning models (LRMs) can automatically allocate more reasoning strengths (\ie the number of reasoning tokens) for harder problems, exhibiting difficulty-awareness for better task performance. While this automatic reasoning strength allocation phenomenon has been widely observed, its underlying mechanism remains largely unexplored. To this end, we provide explanations for this phenomenon from the perspective of model activations. \textbf{We find evidence that LRMs pre-plan the reasoning strengths in their activations even before generation, with this reasoning strength causally controlled by the magnitude of a pre-allocated directional vector. } Specifically, we show that the number of reasoning tokens is predictable solely based on the question activations using linear probes, indicating that LRMs estimate the required reasoning strength in advance. We then uncover that LRMs encode this reasoning strength through a pre-allocated directional vector embedded in the activations of the model, where the vector’s magnitude modulates the reasoning strength. Subtracting this vector can lead to reduced reasoning token number and performance, while adding this vector can lead to increased reasoning token number and even improved performance. We further reveal that this direction vector consistently yields positive reasoning length prediction, and it modifies the logits of end-of-reasoning token \texttt{} to affect the reasoning length. Finally, we demonstrate two potential applications of our findings: overthinking behavior detection and enabling efficient reasoning on simple problems. Our work provides new insights into the internal mechanisms of reasoning in LRMs and offers practical tools for controlling their reasoning behaviors. Our code is available at \url{https: //anonymous. 4open. science/r/LRM-plans-CoT-7E04}.

PDF Details

NeurIPS Conference 2025 Conference Paper

RSafe: Incentivizing proactive reasoning to build robust and adaptive LLM safeguards

jingnan zheng
Xiangtian Ji
Yijun Lu
Chenhang Cui
Weixiang Zhao
Gelei Deng
Zhenkai Liang
An Zhang

Large Language Models (LLMs) continue to exhibit vulnerabilities despite deliberate safety alignment efforts, posing significant risks to users and society. To safeguard against the risk of policy-violating content, system-level moderation via external guard models—designed to monitor LLM inputs and outputs and block potentially harmful content—has emerged as a prevalent mitigation strategy. Existing approaches of training guard models rely heavily on extensive human curated datasets and struggle with out-of-distribution threats, such as emerging harmful categories or jailbreak attacks. To address these limitations, we propose RSafe, an adaptive reasoning-based safeguard that conducts guided safety reasoning to provide robust protection within the scope of specified safety policies. RSafe operates in two stages: (1) guided reasoning, where it analyzes safety risks of input content through policy-guided step-by-step reasoning, and (2) reinforced alignment, where rule-based RL optimizes its reasoning paths to align with accurate safety prediction. This two-stage training paradigm enables RSafe to internalize safety principles to generalize safety protection capability over unseen or adversarial safety violation scenarios. During inference, RSafe accepts user-specified safety policies to provide enhanced safeguards tailored to specific safety requirements. Experiments demonstrate that RSafe matches state-of-the-art guard models using limited amount of public data in both prompt- and response-level harmfulness detection, while achieving superior out-of-distribution generalization on both emerging harmful category and jailbreak attacks. Furthermore, RSafe provides human-readable explanations for its safety judgments for better interpretability. RSafe offers a robust, adaptive, and interpretable solution for LLM safety moderation, advancing the development of reliable safeguards in dynamic real-world environments. Our code is available at https: //anonymous. 4open. science/r/RSafe-996D.

PDF Details

NeurIPS Conference 2025 Conference Paper

Teaching Language Models to Evolve with Users: Dynamic Profile Modeling for Personalized Alignment

Weixiang Zhao
Xingyu Sui
Yulin Hu
Jiahe Guo
Haixiao Liu
Biye Li
Yanyan Zhao
Bing Qin

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.

PDF Details

NeurIPS Conference 2025 Conference Paper

When Less Language is More: Language-Reasoning Disentanglement Makes LLMs Better Multilingual Reasoners

Weixiang Zhao
Jiahe Guo
Yang Deng
Tongtong Wu
Wenxuan Zhang
Yulin Hu
Xingyu Sui
Yanyan Zhao

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.

PDF Details

AAAI Conference 2023 Conference Paper

Knowledge-Bridged Causal Interaction Network for Causal Emotion Entailment

Weixiang Zhao
Yanyan Zhao
Zhuojun Li
Bing Qin

Causal Emotion Entailment aims to identify causal utterances that are responsible for the target utterance with a non-neutral emotion in conversations. Previous works are limited in thorough understanding of the conversational context and accurate reasoning of the emotion cause. To this end, we propose Knowledge-Bridged Causal Interaction Network (KBCIN) with commonsense knowledge (CSK) leveraged as three bridges. Specifically, we construct a conversational graph for each conversation and leverage the event-centered CSK as the semantics-level bridge (S-bridge) to capture the deep inter-utterance dependencies in the conversational context via the CSK-Enhanced Graph Attention module. Moreover, social-interaction CSK serves as emotion-level bridge (E-bridge) and action-level bridge (A-bridge) to connect candidate utterances with the target one, which provides explicit causal clues for the Emotional Interaction module and Actional Interaction module to reason the target emotion. Experimental results show that our model achieves better performance over most baseline models. Our source code is publicly available at https://github.com/circle-hit/KBCIN.

PDF Details DOI

IJCAI Conference 2022 Conference Paper

CauAIN: Causal Aware Interaction Network for Emotion Recognition in Conversations

Weixiang Zhao
Yanyan Zhao
Xin Lu

Emotion Recognition in Conversations has attained increasing interest in the natural language processing community. Many neural-network based approaches endeavor to solve the challenge of emotional dynamics in conversations and gain appealing results. However, these works are limited in capturing deep emotional clues in conversational context because they ignore the emotion cause that could be viewed as stimulus to the target emotion. In this work, we propose Causal Aware Interaction Network (CauAIN) to thoroughly understand the conversational context with the help of emotion cause detection. Specifically, we retrieve causal clues provided by commonsense knowledge to guide the process of causal utterance traceback. Both retrieve and traceback steps are performed from the perspective of intra- and inter-speaker interaction simultaneously. Experimental results on three benchmark datasets show that our model achieves better performance over most baseline models.

PDF Details DOI

AIIM Journal 2011 Journal Article

A modified artificial immune system based pattern recognition approach—An application to clinical diagnostics

Weixiang Zhao
Cristina E. Davis

Objective This paper introduces a modified artificial immune system (AIS)-based pattern recognition method to enhance the recognition ability of the existing conventional AIS-based classification approach and demonstrates the superiority of the proposed new AIS-based method via two case studies of breast cancer diagnosis. Methods and materials Conventionally, the AIS approach is often coupled with the k nearest neighbor (k-NN) algorithm to form a classification method called AIS-kNN. In this paper we discuss the basic principle and possible problems of this conventional approach, and propose a new approach where AIS is integrated with the radial basis function – partial least square regression (AIS-RBFPLS). Additionally, both the two AIS-based approaches are compared with two classical and powerful machine learning methods, back-propagation neural network (BPNN) and orthogonal radial basis function network (Ortho-RBF network). Results The diagnosis results show that: (1) both the AIS-kNN and the AIS-RBFPLS proved to be a good machine leaning method for clinical diagnosis, but the proposed AIS-RBFPLS generated an even lower misclassification ratio, especially in the cases where the conventional AIS-kNN approach generated poor classification results because of possible improper AIS parameters. For example, based upon the AIS memory cells of “replacement threshold=0. 3”, the average misclassification ratios of two approaches for study 1 are 3. 36% (AIS-RBFPLS) and 9. 07% (AIS-kNN), and the misclassification ratios for study 2 are 19. 18% (AIS-RBFPLS) and 28. 36% (AIS-kNN); (2) the proposed AIS-RBFPLS presented its robustness in terms of the AIS-created memory cells, showing a smaller standard deviation of the results from the multiple trials than AIS-kNN. For example, using the result from the first set of AIS memory cells as an example, the standard deviations of the misclassification ratios for study 1 are 0. 45% (AIS-RBFPLS) and 8. 71% (AIS-kNN) and those for study 2 are 0. 49% (AIS-RBFPLS) and 6. 61% (AIS-kNN); and (3) the proposed AIS-RBFPLS classification approaches also yielded better diagnosis results than two classical neural network approaches of BPNN and Ortho-RBF network. Conclusion In summary, this paper proposed a new machine learning method for complex systems by integrating the AIS system with RBFPLS. This new method demonstrates its satisfactory effect on classification accuracy for clinical diagnosis, and also indicates its wide potential applications to other diagnosis and detection problems.

Details DOI