Yixuan Wang

EAAI Journal 2026 Journal Article

An autoregressive framework for reconstructing editable parametric computer-aided design models from point clouds

• Jiaxing Lu
• Yixuan Wang
• Yangyong Wu
• Heran Li
• Yan Shi
• Fangwei Ning

AAAI Conference 2026 Conference Paper

CAMERA: Multi-Matrix Joint Compression for MoE Models via Micro-Expert Redundancy Analysis

• Yuzhuang Xu
• Xu Han
• Yuanchi Zhang
• Yixuan Wang
• Yijun Liu
• Shiyu Ji
• Qingfu Zhu
• Wanxiang Che

Large Language Models (LLMs) with Mixture-of-Experts (MoE) architectures are distinguished by their strong performance scaling with increasing parameters across a wide range of tasks, yet they also suffer from substantial computational and storage overheads. Notably, the performance gains of MoE models do not scale proportionally with the growth in expert parameters. While prior works attempt to reduce parameters via expert-level pruning, merging, or decomposition, they still suffer from challenges in both performance and computational efficiency. In this paper, we address these challenges by introducing micro-expert as a finer-grained compression unit that spans across matrices. We first establish a more fundamental perspective, viewing MoE layers as mixtures of micro-experts, and present CAMERA, a lightweight and training-free framework for identifying micro-expert redundancy. Our analysis uncovers significant variance in micro-expert contributions during decoding. Based on this insight, we further propose CAMERA-P, a structured micro-expert pruning framework, and CAMERA-Q, a mixed-precision quantization idea designed for micro-experts. Extensive experiments on nine downstream tasks show that CAMERA-P consistently outperforms strong baselines under pruning ratios ranging from 20% to 60%. Furthermore, CAMERA-Q achieves superior results under aggressive 2-bit quantization, surpassing existing matrix- and channel-level ideas. Notably, our method enables complete micro-expert analysis of Qwen2-57B-A14B in less than 5 minutes on a single NVIDIA A100-40GB GPU.

AAAI Conference 2026 Conference Paper

Judge Q: Trainable Queries for Optimized Information Retention in KV Cache Eviction

• Yijun Liu
• Yixuan Wang
• Yuzhuang Xu
• Shiyu Ji
• Yang Xu
• Qingfu Zhu
• Wanxiang Che

Large language models (LLMs) utilize key-value (KV) cache to store historical information during sequence processing. The size of KV cache grows linearly as the length of the sequence extends, which seriously affects memory usage and decoding efficiency. Current methods for KV cache eviction typically utilize the last window from the pre-filling phase as queries to compute the KV importance scores for eviction. Although this scheme is simple to implement, it tends to overly focus on local information, potentially leading to the neglect or omission of crucial global information. To mitigate this issue, we propose **Judge Q**, a novel training method which incorporates a soft token list. This method only tunes the model’s embedding layer at a low training cost. By concatenating the soft token list at the end of the input sequence, we train these tokens' attention map to the original input sequence to align with that of the actual decoded tokens. In this way, the queries corresponding to the soft tokens can effectively capture global information and better evaluate the importance of the keys and values within the KV cache, thus maintaining decoding quality when KV cache is evicted. Under the same eviction budget, our method exhibits less performance degradation compared to existing eviction approaches. We validate our approach through experiments conducted on models such as Llama-3.1-8B-Instruct and Mistral-7B-Instruct-v0.3, using benchmarks including LongBench, RULER, and Needle-in-a-Haystack. Results indicate an improvement of approximately 1 point on the LongBench and over 3 points on RULER. This proposed methodology can be seamlessly integrated into existing open-source models with minimal training overhead, thereby enhancing performance in KV cache eviction scenarios.

ECAI Conference 2025 Conference Paper

A Style-Aware Polytomous Diagnostic Model for Individual Traits

• Yixuan Wang
• Jiale Feng
• Yue Huang
• Xuruo Pan
• Zhongjing Huang
• Zhi Liu
• Hong Qian

Diagnostic models aim to precisely infer individuals’ cognitive or non-cognitive competencies from their response logs, such as mathematical or social-emotional skills. While deep learning shows success in cognitive diagnosis, it remains underexplored in the equally important area of non-cognitive trait diagnosis. Accurate non-cognitive trait estimation is critical for individuals’ development. Unlike cognitive assessments using right or wrong responses, non-cognitive trait assessments typically use subjective Likert-scale items with ordinal polytomous options to reflect latent trait levels. Furthermore, individual response styles, such as tendencies toward higher or lower options, introduce bias in trait inference, causing estimations that deviate from true trait levels. Thus, maintaining options ordinal semantic structure and mitigating the response style bias in trait estimation are two major challenges for accurate trait diagnosis. To address these issues, this paper proposes a Style-Aware Polytomous Diagnosis (SAPD) model. Specifically, to capture the ordinal semantics of response options, SAPD constructs an Ordinal Option Graph (OOG) that explicitly encodes the ordinal relationship among polytomous options, where higher options reflect higher latent trait levels. To mitigate the bias caused by individual response styles, we first design a Style-Aware Relational Graph (SARG), a heterogeneous graph that integrates multiple interactions among participants, items, options and traits, implicitly embedding response style information within node representations. We then propose a Response Style Corrector (RSC) that explicitly captures individual response tendencies and disentangles response style bias during trait diagnosis, allowing for dynamic and adaptive correction of trait levels. Extensive experiments on five real-world datasets show that SAPD improves accuracy by an average of 4% over competitive methods. Visualizations confirm SAPD effectively disentangles response style effects, leading to more accurate and interpretable trait diagnosis.

ICLR Conference 2025 Conference Paper

KAN: Kolmogorov-Arnold Networks

• Ziming Liu 0001
• Yixuan Wang
• Sachin Vaidya
• Fabian Ruehle
• James Halverson
• Marin Soljacic
• Thomas Y. Hou 0001
• Max Tegmark

Inspired by the Kolmogorov-Arnold representation theorem, we propose Kolmogorov-Arnold Networks (KANs) as promising alternatives to Multi-Layer Perceptrons (MLPs). While MLPs have fixed activation functions on nodes ("neurons''), KANs have learnable activation functions on edges ("weights''). KANs have no linear weights at all -- every weight parameter is replaced by a univariate function parametrized as a spline. We show that this seemingly simple change makes KANs outperform MLPs in terms of accuracy and interpretability, on small-scale AI + Science tasks. For accuracy, smaller KANs can achieve comparable or better accuracy than larger MLPs in function fitting tasks. Theoretically and empirically, KANs possess faster neural scaling laws than MLPs. For interpretability, KANs can be intuitively visualized and can easily interact with human users. Through two examples in mathematics and physics, KANs are shown to be useful ``collaborators'' helping scientists (re)discover mathematical and physical laws. In summary, KANs are promising alternatives for MLPs. Despite the slow training of KANs, their improved accuracy and interpretability show the potential to improve today's deep learning models which rely heavily on MLPs. More research is necessary to make KANs' training more efficient.

NeurIPS Conference 2025 Conference Paper

Long-term Intracortical Neural activity and Kinematics (LINK): An intracortical neural dataset for chronic brain-machine interfaces, neuroscience, and machine learning

• Hisham Temmar
• Yixuan Wang
• Nina Gill
• Nicholas Mellon
• Chang Liu
• Luis Cubillos
• Rio Parsons
• Joseph Costello

Intracortical brain-machine interfaces (iBMIs) have enabled movement and speech in people living with paralysis by using neural data to decode behaviors in real-time. However, intracortical neural recordings exhibit significant instabilities over time, which poses problems for iBMIs, neuroscience, and machine learning. For iBMIs, neural instabilities require frequent decoder recalibration to maintain high performance, a critical bottleneck for real-world translation. Several approaches have been developed to address this issue, and the field has recognized the need for standardized datasets on which to compare them, but no standard dataset exists for evaluation over year-long timescales. In neuroscience, a growing body of research attempts to elucidate the latent computations performed by populations of neurons. Nonstationarity in neural recordings imposes significant challenges to the design of these studies, so a dataset containing recordings over large time spans would improve methods to account for instabilities. In machine learning, continuous domain adaptation of temporal data is an area of active research, and a dataset containing shift distributions on long time scales would be beneficial to researchers. To address these gaps, we present the LINK Dataset (Long-term Intracortical Neural activity and Kinematics), which contains intracortical spiking activity and kinematic data from 312 sessions of a non-human primate performing a dexterous, 2 degree-of-freedom finger movement task, spanning 1, 242 days. We also present longitudinal analyses of the dataset’s neural spiking activity and its relationship to kinematics, as well as overall decoding performance using linear and neural network models. The LINK dataset (https: //dandiarchive. org/dandiset/001201) and code (https: //github. com/chesteklab/LINK_dataset) are freely available to the public.

EAAI Journal 2025 Journal Article

Neural network adaptive force control for pneumatic polishing end-actuator with external disturbances and full-state constrains

• Zhiguo Yang
• Jiange Kou
• Zhanxin Li
• Wenbo Zhao
• Yushan Ma
• Yixuan Wang
• Yan Shi

ICLR Conference 2025 Conference Paper

On the expressiveness and spectral bias of KANs

• Yixuan Wang
• Jonathan W. Siegel
• Ziming Liu 0001
• Thomas Y. Hou 0001

Kolmogorov-Arnold Networks (KAN) \cite{liu2024kan} were very recently proposed as a potential alternative to the prevalent architectural backbone of many deep learning models, the multi-layer perceptron (MLP). KANs have seen success in various tasks of AI for science, with their empirical efficiency and accuracy demonstrated in function regression, PDE solving, and many more scientific problems. In this article, we revisit the comparison of KANs and MLPs, with emphasis on a theoretical perspective. On the one hand, we compare the representation and approximation capabilities of KANs and MLPs. We establish that MLPs can be represented using KANs of a comparable size. This shows that the approximation and representation capabilities of KANs are at least as good as MLPs. Conversely, we show that KANs can be represented using MLPs, but that in this representation the number of parameters increases by a factor of the KAN grid size. This suggests that KANs with a large grid size may be more efficient than MLPs at approximating certain functions. On the other hand, from the perspective of learning and optimization, we study the spectral bias of KANs compared with MLPs. We demonstrate that KANs are less biased toward low frequencies than MLPs. We highlight that the multi-level learning feature specific to KANs, i.e. grid extension of splines, improves the learning process for high-frequency components. Detailed comparisons with different choices of depth, width, and grid sizes of KANs are made, shedding some light on how to choose the hyperparameters in practice.

NeurIPS Conference 2025 Conference Paper

Robustifying Learning-Augmented Caching Efficiently without Compromising 1-Consistency

• Peng Chen
• Hailiang Zhao
• Jiaji Zhang
• Xueyan Tang
• Yixuan Wang
• Shuiguang Deng

The online caching problem aims to minimize cache misses when serving a sequence of requests under a limited cache size. While naive learning-augmented caching algorithms achieve ideal $1$-consistency, they lack robustness guarantees. Existing robustification methods either sacrifice $1$-consistency or introduce excessive computational overhead. In this paper, we introduce Guard, a lightweight robustification framework that enhances the robustness of a broad class of learning-augmented caching algorithms to $2H_{k-1} + 2$, while preserving their $1$-consistency. Guard achieves the current best-known trade-off between consistency and robustness, with only $\mathcal{O}(1)$ additional per-request overhead, thereby maintaining the original time complexity of the base algorithm. Extensive experiments across multiple real-world datasets and prediction models validate the effectiveness of Guard in practice.

NeurIPS Conference 2024 Conference Paper

HORSE: Hierarchical Representation for Large-Scale Neural Subset Selection

• Binghui Xie
• Yixuan Wang
• Yongqiang Chen
• Kaiwen Zhou
• Yu Li
• Wei Meng
• James Cheng

Subset selection tasks, such as anomaly detection and compound selection in AI-assisted drug discovery, are crucial for a wide range of applications. Learning subset-valued functions with neural networks has achieved great success by incorporating permutation invariance symmetry into the architecture. However, existing neural set architectures often struggle to either capture comprehensive information from the superset or address complex interactions within the input. Additionally, they often fail to perform in scenarios where superset sizes surpass available memory capacity. To address these challenges, we introduce the novel concept of the Identity Property, which requires models to integrate information from the originating set, resulting in the development of neural networks that excel at performing effective subset selection from large supersets. Moreover, we present the Hierarchical Representation of Neural Subset Selection (HORSE), an attention-based method that learns complex interactions and retains information from both the input set and the optimal subset supervision signal. Specifically, HORSE enables the partitioning of the input ground set into manageable chunks that can be processed independently and then aggregated, ensuring consistent outcomes across different partitions. Through extensive experimentation, we demonstrate that HORSE significantly enhances neural subset selection performance by capturing more complex information and surpasses state-of-the-art methods in handling large-scale inputs by a margin of up to 20%.

NeurIPS Conference 2024 Conference Paper

Off-Dynamics Reinforcement Learning via Domain Adaptation and Reward Augmented Imitation

• Yihong Guo
• Yixuan Wang
• Yuanyuan Shi
• Pan Xu
• Anqi Liu

Training a policy in a source domain for deployment in the target domain under a dynamics shift can be challenging, often resulting in performance degradation. Previous work tackles this challenge by training on the source domain with modified rewards derived by matching distributions between the source and the target optimal trajectories. However, pure modified rewards only ensure the behavior of the learned policy in the source domain resembles trajectories produced by the target optimal policies, which does not guarantee optimal performance when the learned policy is actually deployed to the target domain. In this work, we propose to utilize imitation learning to transfer the policy learned from the reward modification to the target domain so that the new policy can generate the same trajectories in the target domain. Our approach, Domain Adaptation and Reward Augmented Imitation Learning (DARAIL), utilizes the reward modification for domain adaptation and follows the general framework of generative adversarial imitation learning from observation (GAIfO) by applying a reward augmented estimator for the policy optimization step. Theoretically, we present an error bound for our method under a mild assumption regarding the dynamics shift to justify the motivation of our method. Empirically, our method outperforms the pure modified reward method without imitation learning and also outperforms other baselines in benchmark off-dynamics environments.

ICRA Conference 2024 Conference Paper

Open X-Embodiment: Robotic Learning Datasets and RT-X Models: Open X-Embodiment Collaboration

• Abby O'Neill
• Abdul Rehman
• Abhiram Maddukuri
• Abhishek Gupta 0004
• Abhishek Padalkar
• Abraham Lee
• Acorn Pooley
• Agrim Gupta

Large, high-capacity models trained on diverse datasets have shown remarkable successes on efficiently tackling downstream applications. In domains from NLP to Computer Vision, this has led to a consolidation of pretrained models, with general pretrained backbones serving as a starting point for many applications. Can such a consolidation happen in robotics? Conventionally, robotic learning methods train a separate model for every application, every robot, and even every environment. Can we instead train "generalist" X-robot policy that can be adapted efficiently to new robots, tasks, and environments? In this paper, we provide datasets in standardized data formats and models to make it possible to explore this possibility in the context of robotic manipulation, alongside experimental results that provide an example of effective X-robot policies. We assemble a dataset from 22 different robots collected through a collaboration between 21 institutions, demonstrating 527 skills (160266 tasks). We show that a high-capacity model trained on this data, which we call RT-X, exhibits positive transfer and improves the capabilities of multiple robots by leveraging experience from other platforms. The project website is robotics-transformer-x. github.io.

AAAI Conference 2024 Conference Paper

REGLO: Provable Neural Network Repair for Global Robustness Properties

• Feisi Fu
• Zhilu Wang
• Weichao Zhou
• Yixuan Wang
• Jiameng Fan
• Chao Huang
• Qi Zhu
• Xin Chen

We present REGLO, a novel methodology for repairing pretrained neural networks to satisfy global robustness and individual fairness properties. A neural network is said to be globally robust with respect to a given input region if and only if all the input points in the region are locally robust. This notion of global robustness also captures the notion of individual fairness as a special case. We prove that any counterexample to a global robustness property must exhibit a corresponding large gradient. For ReLU networks, this result allows us to efficiently identify the linear regions that violate a given global robustness property. By formulating and solving a suitable robust convex optimization problem, REGLO then computes a minimal weight change that will provably repair these violating linear regions.

JAIR Journal 2024 Journal Article

Towards Trustworthy AI-Enabled Decision Support Systems: Validation of the Multisource AI Scorecard Table (MAST)

• Pouria Salehi
• Yang Ba
• Nayoung Kim
• Ahmadreza Mosallanezhad
• Anna Pan
• Myke C. Cohen
• Yixuan Wang
• Jieqiong Zhao

The Multisource AI Scorecard Table (MAST) is a checklist tool to inform the design and evaluation of trustworthy AI systems based on the U.S. Intelligence Community’s analytic tradecraft standards. In this study, we investigate whether MAST can be used to differentiate between high and low trustworthy AI-enabled decision support systems (AI-DSSs). Evaluating trust in AI-DSSs poses challenges to researchers and practitioners. These challenges include identifying the components, capabilities, and potential of these systems, many of which are based on the complex deep learning algorithms that drive DSS performance and preclude complete manual inspection. Using MAST, we developed two interactive AI-DSS testbeds. One emulated an identity-verification task in security screening, and another emulated a text-summarization system to aid in an investigative task. Each testbed had one version designed to reach low MAST ratings, and another designed to reach high MAST ratings. We hypothesized that MAST ratings would be positively related to the trust ratings of these systems. A total of 177 subject-matter experts were recruited to interact with and evaluate these systems. Results generally show higher MAST ratings for the high-MAST compared to the low-MAST groups, and that measures of trust perception are highly correlated with the MAST ratings. We conclude that MAST can be a useful tool for designing and evaluating systems that will engender trust perceptions, including for AI-DSS that may be used to support visual screening or text summarization tasks. However, higher MAST ratings may not translate to higher joint performance, and the connection between MAST and appropriate trust or trustworthiness remains an open question.

NeurIPS Conference 2024 Conference Paper

Variational Delayed Policy Optimization

• Qingyuan Wu
• Simon S. Zhan
• Yixuan Wang
• Yuhui Wang
• Chung-Wei Lin
• Chen Lv
• Qi Zhu
• Chao Huang

In environments with delayed observation, state augmentation by including actions within the delay window is adopted to retrieve Markovian property to enable reinforcement learning (RL). Whereas, state-of-the-art (SOTA) RL techniques with Temporal-Difference (TD) learning frameworks commonly suffer from learning inefficiency, due to the significant expansion of the augmented state space with the delay. To improve the learning efficiency without sacrificing performance, this work novelly introduces Variational Delayed Policy Optimization (VDPO), reforming delayed RL as a variational inference problem. This problem is further modelled as a two-step iterative optimization problem, where the first step is TD learning in the delay-free environment with a small state space, and the second step is behaviour cloning which can be addressed much more efficiently than TD learning. We not only provide a theoretical analysis of VDPO in terms of sample complexity and performance, but also empirically demonstrate that VDPO can achieve consistent performance with SOTA methods, with a significant enhancement of sample efficiency (approximately 50\% less amount of samples) in the MuJoCo benchmark.

JBHI Journal 2023 Journal Article

Cross-Hospital Sepsis Early Detection via Semi-Supervised Optimal Transport With Self-Paced Ensemble

• Ruiqing Ding
• Yu Zhou
• Jie Xu
• Yan Xie
• Qiqiang Liang
• He Ren
• Yixuan Wang
• Yanlin Chen

Leveraging machine learning techniques for Sepsis early detection and diagnosis has attracted increasing interest in recent years. However, most existing methods require a large amount of labeled training data, which may not be available for a target hospital that deploys a new Sepsis detection system. More seriously, as treated patients are diversified between hospitals, directly applying a model trained on other hospitals may not achieve good performance for the target hospital. To address this issue, we propose a novel semi-supervised transfer learning framework based on optimal transport theory and self-paced ensemble for Sepsis early detection, called SPSSOT, which can efficiently transfer knowledge from the source hospital (with rich labeled data) to the target hospital (with scarce labeled data). Specifically, SPSSOT incorporates a new optimal transport-based semi-supervised domain adaptation component that can effectively exploit all the unlabeled data in the target hospital. Moreover, self-paced ensemble is adapted in SPSSOT to alleviate the class imbalance issue during transfer learning. In a nutshell, SPSSOT is an end-to-end transfer learning method that automatically selects suitable samples from two domains (hospitals) respectively and aligns their feature spaces. Extensive experiments on two open clinical datasets, MIMIC-III and Challenge, demonstrate that SPSSOT outperforms state-of-the-art transfer learning methods by improving 1–3% of AUC.