Author name cluster

Yu Huang

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.

28 papers

1 author row

AAAI Conference 2026 Conference Paper

Beyond Conservation: Flexible Molecular Assembly with Unbalanced Diffusion Bridge

Rongchao Zhang
Yiwei Lou
Yu Huang
Yi Xin
Yongzhi Cao
Hanpin Wang

Molecular assembly (MA) has long been a fundamental task in chemistry and biology, with the potential to create new materials and enable novel functions beyond the molecular scale. However, its vast conformational search space poses substantial challenges, and current generative models remain limited in capturing molecular flexibility and preventing non-physical poses. In this paper, we propose AssemUDB, a diffusion bridge–based framework that learns transport mappings between two distinct flexible domains for molecular assembly generation. We reformulate the marginal matching constraint of diffusion bridges as a coupling distribution governed by unbalanced transport rather than imposing strict conservation. Subsequently, we employ a progressive process from structural relaxation in Euclidean space to assembly on the SE(3) manifold. This relaxation of marginal conservation grants the generative model greater flexibility and leads to more physically plausible atom placements. Comprehensive experiments demonstrate the superior performance of AssemUDB. Notably, we find that the method demonstrates performance comparable to, or even better than, mature tools such as PackMol for packing tasks.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Efficient Segmentation with Multimodal Large Language Model via Token Routing

Changsong Wen
Zelin Peng
Yu Huang
Wei Shen

Recent advances in multimodal large language models (MLLMs) have demonstrated strong capabilities in addressing open-world segmentation tasks. However, the substantial computational cost of the LLM components presents a significant challenge, especially in segmentation tasks, where efficiency has long been a central concern. Existing efficient MLLM approaches typically reduce computation cost by pruning visual tokens in the early layers, as they account for the majority of the input sequence. Despite their efficiency, this is incompatible with dense prediction tasks such as segmentation, since removing visual tokens leads to the loss of essential object parts and spatial details. To better understand the roles of visual tokens in segmentation, we analyze the attention weights of both image and mask tokens within LLM. We find that image tokens are important throughout all layers, whereas mask tokens only attend to image tokens at deeper layers. Based on the observation, we build an efficient segmentation framework based on MLLMs by introducing a sophisticated token routing strategy. This strategy dynamically determines when and how different tokens participate in computation: For mask tokens, they are only inserted at deeper layers of the LLM to reduce redundant computation, since they rarely attend to image tokens in early layers; For image tokens, only a small number of them, named proxies, are updated via full feedforward network (FFN) computation, while the update of the remaining tokens is guided by these proxies, i.e., efficiently computed through a lightweight projector applied on the difference of the proxies during their update. Our method achieves a 1.5× acceleration over the original LLM process by reducing its FLOPs to 56%, while maintaining the same segmentation performance.

PDF Details DOI

AAAI Conference 2026 Conference Paper

MedMKEB: A Comprehensive Knowledge Editing Benchmark for Medical Multimodal Large Language Models

Dexuan Xu
Jieyi Wang
Zhongyan Chai
Yongzhi Cao
Hanpin Wang
Huamin Zhang
Yu Huang

Recent advances in multimodal large language models (MLLMs) have significantly improved medical AI, enabling it to unify the understanding of visual and textual information. However, as medical knowledge continues to evolve, it is critical to allow these models to efficiently update outdated or incorrect information without retraining from scratch. Although textual knowledge editing has been widely studied, there is still a lack of systematic benchmarks for multimodal medical knowledge editing involving image and text modalities. To fill this gap, we present MedMKEB, the first comprehensive benchmark designed to evaluate the reliability, generality, locality, portability, and robustness of knowledge editing in medical multimodal large language models. MedMKEB is built on a high-quality medical visual question-answering dataset and enriched with carefully constructed editing tasks, including counterfactual correction, semantic generalization, knowledge transfer, and adversarial robustness. We incorporate human expert validation to ensure the accuracy and reliability of the benchmark. Extensive experiments on state-of-the-art general and medical MLLMs demonstrate the limitations of existing knowledge editing methods in the medical domain, highlighting the need to develop specialized editing strategies.

PDF Details DOI

TMLR Journal 2026 Journal Article

Multimodal Deception in Explainable AI: Concept-Level Backdoor Attacks on Concept Bottleneck Models

Songning Lai
Jiayu Yang
Yu Huang
Lijie Hu
TianlangXue
Zhangyi Hu
Jiaxu Li
Haicheng Liao

Deep learning has demonstrated transformative potential across domains, yet its inherent opacity has driven the development of Explainable Artificial Intelligence (XAI). Concept Bottleneck Models (CBMs), which enforce interpretability through human-understandable concepts, represent a prominent advancement in XAI. However, despite their semantic transparency, CBMs remain vulnerable to security threats such as backdoor attacks—malicious manipulations that induce controlled misbehaviors during inference. While CBMs leverage multimodal representations (visual inputs and textual concepts) to enhance interpretability, their dual-modality structure introduces unique, unexplored attack surfaces. To address this risk, we propose CAT (Concept-level Backdoor ATtacks), a methodology that injects stealthy triggers into conceptual representations during training. Unlike naive attacks that randomly corrupt concepts, CAT employs a sophisticated filtering mechanism to enable precise prediction manipulation without compromising clean-data performance. We further propose CAT+, an enhanced variant incorporating a concept correlation function to iteratively optimize trigger-concept associations, thereby maximizing attack effectiveness and stealthiness. Crucially, we validate our approach through a rigorous two-stage evaluation framework. First, we establish the fundamental vulnerability of the concept bottleneck layer in a controlled setting, showing that CAT+ achieves high attack success rates (ASR) while remaining statistically indistinguishable from natural data. Second, we demonstrate practical end-to-end feasibility via our proposed Image2Trigger_c method, which translates visual perturbations into concept-level triggers, achieving an end-to-end ASR of 53.29%. Extensive experiments show that CAT outperforms random-selection baselines significantly, and standard defenses like Neural Cleanse fail to detect these semantic attacks. This work highlights critical security risks in interpretable AI systems and provides a robust methodology for future security assessments of CBMs.

PDF Details

AAAI Conference 2026 Conference Paper

Multitasks-based Deep Evidential Fusion Network for Blind Image Quality Assessment

Yiwei Lou
Yuanpeng He
Rongchao Zhang
Yongzhi Cao
Hanpin Wang
Yu Huang

Blind image quality assessment (BIQA) methods often incorporate auxiliary tasks to improve performance. However, existing approaches face limitations due to insufficient integration and a lack of flexible uncertainty estimation, leading to suboptimal performance. To address these challenges, we propose a multitasks-based Deep Evidential Fusion Network (DEFNet) for BIQA, which performs multitask optimization with the assistance of scene and distortion type classification tasks. To achieve a more robust and reliable representation, we design a novel trustworthy information fusion strategy. It first combines diverse features and patterns across sub-regions to enhance information richness, and then performs local-global information fusion by balancing fine-grained details with coarse-grained context. Moreover, DEFNet exploits advanced uncertainty estimation technique inspired by evidential learning with the help of normal-inverse gamma distribution mixture. Extensive experiments on both synthetic and authentic distortion datasets demonstrate the effectiveness and robustness of the proposed framework. Additional evaluation and analysis are carried out to highlight its strong generalization capability and adaptability to previously unseen scenarios.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Scene Experts: Specializing in 3D Gaussian Splatting with Adaptive Decomposition

Xiaowen Fu
Yang Zhang
Yuhan Tang
Huazhong Zhang
Tianxing Zhao
Yuhang Guo
Yu Huang
Jinbao Wang

Anchor-based 3D Gaussian Splatting (GS), exemplified by Scaffold-GS, achieves remarkable storage efficiency through a hybrid explicit-implicit representation. However, their reliance on a single, monolithic network to decode anchor features imposes a severe bottleneck on model capacity, often resulting in blurred details and view-dependent artifacts in complex scenes. To break this bottleneck, we introduce the concept of Scene Experts: a strategy that decomposes the task of modeling a complex scene across a collection of specialized sub-models. To realize the paradigm, we propose MoE-GS. Our approach designs the decoder as a Sparsely-Gated Mixture of Experts (MoE), which dramatically increases the model's total capacity while maintaining comparable inference cost via sparse activation. To effectively train this high-capacity model, we propose two key innovations: (1) A progressive curriculum learning strategy that first trains all experts on a robust baseline before encouraging them to specialize on different scene components. (2) A novel opacity-aware regularization that penalizes inactive neural Gaussians, ensuring the expanded capacity is efficiently used. Extensive experiments demonstrate that MoE-GS substantially outperforms state-of-the-art methods on diverse benchmarks, significantly improving reconstruction fidelity while requiring a smaller or comparable Gaussian model size.

PDF Details DOI

AAAI Conference 2026 Conference Paper

VisionReward: Fine-Grained Multi-Dimensional Human Preference Learning for Image and Video Generation

Jiazheng Xu
Yu Huang
Jiale Cheng
Yuanming Yang
Jiajun Xu
Yuan Wang
Wenbo Duan
Shen Yang

Visual generative models have achieved remarkable progress in synthesizing photorealistic images and videos, yet aligning their outputs with human preferences across critical dimensions remains a persistent challenge. Though reinforcement learning from human feedback offers promise for preference alignment, existing reward models for visual generation face limitations, including black-box scoring without interpretability and potentially resultant unexpected biases. We present VisionReward, a general framework for learning human visual preferences in both image and video generation. Specifically, we employ a hierarchical visual assessment framework to capture fine-grained human preferences, and leverages linear weighting to enable interpretable preference learning. Furthermore, we propose a multi-dimensional consistent strategy when using VisionReward as a reward model during preference optimization for visual generation. Experiments show that VisionReward can significantly outperform existing image and video reward models on both machine metrics and human evaluation. Notably, VisionReward surpasses VideoScore by 17.2% in preference prediction accuracy, and text-to-video models with VisionReward achieve a 31.6% higher pairwise win rate compared to the same models using VideoScore.

PDF Details DOI

AAAI Conference 2026 Conference Paper

XLinear: A Lightweight and Accurate MLP-Based Model for Long-Term Time Series Forecasting with Exogenous Inputs

Xinyang Chen
Huidong Jin
Yu Huang
Zaiwen Feng

Despite the prevalent assumption of uniform variable importance in long-term time series forecasting models, real-world applications often exhibit asymmetric causal relationships and varying data acquisition costs. Specifically, cost‐effective exogenous data (e.g., local weather) can unilaterally influence dynamics of endogenous variables, such as lake surface temperature. Exploiting these links enables more effective forecasts when exogenous inputs are readily available. Transformer-based models capture long-range dependencies but incur high computation and suffer from permutation invariance. Patch-based variants improve efficiency yet can miss local temporal patterns. To efficiently exploit informative signals across both the temporal dimension and relevant exogenous variables, this study proposes XLinear, a lightweight time series forecasting model built upon Multi-Layer Perceptrons (MLPs). XLinear uses a global token derived from an endogenous variable as a pivotal hub for interacting with exogenous variables, and employs MLPs with sigmoid activation to extract both temporal patterns and variate-wise dependencies. Its prediction head then integrates these signals to forecast the endogenous series. We evaluate XLinear on seven standard benchmarks and five real-world datasets with exogenous inputs. Compared with state-of-the-art models, XLinear delivers superior accuracy and efficiency for both multivariate forecasts and univariate forecasts influenced by exogenous inputs.

PDF Details DOI

YNICL Journal 2025 Journal Article

Alterations of long-range association fibers in patients with anti-N-methyl-D-aspartate receptor encephalitis

Xiaodong Chen
Ling Fang
Yiying Huang
Yu Huang
Yi Lu
Jinhui Wang
Chunxin Liu
Huanquan Liao

BACKGROUND: Patients with anti-NMDAR encephalitis typically exhibit impaired cognitive integration, which relies on the integrity of long-range association fibers connecting diverse brain regions. However, the microstructural integrity of long-range association fibers in this population remains unknown. METHODS: Diffusion tensor imaging (DTI) data were collected from 32 patients with anti-NMDAR encephalitis and 30 healthy controls. Patients were further categorized into early and delayed immunotherapy subgroups based on a 2-week threshold for immunotherapy initiation. The diffusion properties of major long-range association fibers were quantified at both the bundle and node levels. RESULTS: Compared with healthy controls, patients exhibited widespread microstructural damage within long-range association fibers, with more severe alterations in the delayed immunotherapy subgroup (FDR-corrected p < 0.05). In this subgroup(n = 14), radial diffusivity (RD) of left inferior fronto-occipital fasciculus (IFOF), left inferior longitudinal fasciculus (ILF), left superior longitudinal fascicles (SLF), and bilateral arcuate fascicles correlated significantly with global cognition (MMSE, FDR-corrected p < 0.05). Notably, RD also strongly correlated with working memory in the delayed immunotherapy subgroup, showing bundle-wise associations for IFOF (left: r = -0.8315, p = 0.0112; right: r = -0.7044, p = 0.0295), ILF (left: r = -0.7473, p = 0.0243), SLF (left: r = -0.7562, p = 0.0243; right: r = -0.6599, p = 0.0391), and arcuate fasciculus (left: r = -0.7240, p = 0.0272; right: r = -0.6835, p = 0.0333), with left-hemisphere predominance confirmed by node-wise analyses of IFOF, ILF, SLF, and arcuate fasciculus (FDR-corrected p < 0.05). CONCLUSIONS: Our findings highlight widespread microstructural damage in long-range association fibers in patients with anti-NMDAR encephalitis, particularly in those with delayed immunotherapy. This damage may serve as the neurophysiological basis for cognitive impairments, with working memory being most affected.

Details DOI

JMLR Journal 2025 Journal Article

ClimSim-Online: A Large Multi-Scale Dataset and Framework for Hybrid Physics-ML Climate Emulation

Sungduk Yu
Zeyuan Hu
Akshay Subramaniam
Walter Hannah
Liran Peng
Jerry Lin
Mohamed Aziz Bhouri
Ritwik Gupta

Modern climate projections lack adequate spatial and temporal resolution due to computational constraints, leading to inaccuracies in representing critical processes like thunderstorms that occur on the sub-resolution scale. Hybrid methods combining physics with machine learning (ML) offer faster, higher fidelity climate simulations by outsourcing compute-hungry, high-resolution simulations to ML emulators. However, these hybrid physics-ML simulations require domain-specific data and workflows that have been inaccessible to many ML experts. This paper is an extended version of our NeurIPS award-winning ClimSim dataset paper. The ClimSim dataset includes 5.7 billion pairs of multivariate input/output vectors spanning ten years at high temporal resolution, capturing the influence of high-resolution, high-fidelity physics on a host climate simulator's macro-scale state. In this extended version, we introduce a significant new contribution in Section 5, which provides a cross-platform, containerized pipeline to integrate ML models into operational climate simulators for hybrid testing. We also implement various baselines of ML models and hybrid simulators to highlight the ML challenges of building stable, skillful emulators. The data (https://huggingface.co/datasets/LEAP/ClimSim_high-res, also in a low-resolution version at https://huggingface.co/datasets/LEAP/ClimSim_low-res and an aquaplanet version at https://huggingface.co/datasets/LEAP/ClimSim_low-res_aqua-planet) and code (https://leap-stc.github.io/ClimSim and https://github.com/leap-stc/climsim-online) are publicly released to support the development of hybrid physics-ML and high-fidelity climate simulations. [abs] [ pdf ][ bib ] [ code ] &copy JMLR 2025. ( edit, beta )

PDF Details

AAAI Conference 2025 Conference Paper

Exploit Your Latents: Coarse-Grained Protein Backmapping with Latent Diffusion Models

Rongchao Zhang
Yu Huang
Yiwei Lou
Yi Xin
Haixu Chen
Yongzhi Cao
Hanpin Wang

Coarse-grained (CG) molecular dynamics of proteins is a preferred approach to studying large molecules on extended time scales by condensing the entire atomic model into a limited number of pseudo-atoms and preserving the thermodynamic properties of the system. However, the significantly increased efficiency impedes the analysis of substantial physicochemical information, since high-resolution atomic details are sacrificed to accelerate simulation. In this paper, we propose LatCPB, a generative approach based on diffusion that enables high-resolution backmapping of CG proteins. Specifically, our model encodes an all-atom into discrete latent embeddings, aligned with learnable multimodal discrete priors for circumventing posterior collapse and maintaining the discrete properties of the protein sequence. During the generation, we further design a latent diffusion process within the continuous latent space due to the potential stochastics in the data. Moreover, LatCPB performs a contrastive learning strategy in latent space to separate feature representations of various molecules and conformations of the same molecule, thus enhancing the comprehension of molecular representational diversity. Experimental results demonstrate that LatCPB is able to backmap CG proteins effectively and achieve outstanding performance.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

MoleBridge: Synthetic Space Projecting with Discrete Markov Bridges

Rongchao Zhang
Yu Huang
Yongzhi Cao
Hanpin Wang

Molecular synthetic space projecting is a critical technique in de novo molecular design, which aims to rectify molecules without synthesizability guarantee by converting them into synthetic postfix notations. However, the vast synthesizable chemical space and the discrete data modalities involved pose significant challenges to postfix notation conversion benchmarking. In this paper, we exploit conditional probability transitions in discrete state space and introduce MoleBridge, a deep generative model built on the Markov bridge approach for designing postfix notations of molecular synthesis pathways. MoleBridge consists of two iterative optimizations: i) Autoregressive extending of notation tokens from molecular graphs, and ii) generation of discrete reaction postfix notations through Markov bridge, where noisy token blocks are progressively denoised over multi-step iterations. For the challenging second iteration, which demands sensitivity to incorrect generative probability paths within intricate chemical spaces, we employ a thinking and denoising separation approach to denoise. Empirically, we find that MoleBridge is capable of accurately predicting synthesis pathways while exhibiting excellent performance in a variety of application scenarios.

PDF Details

NeurIPS Conference 2025 Conference Paper

Multi-head Transformers Provably Learn Symbolic Multi-step Reasoning via Gradient Descent

Tong Yang
Yu Huang
Yingbin Liang
Yuejie Chi

Transformers have demonstrated remarkable capabilities in multi-step reasoning tasks. However, understandings of the underlying mechanisms by which they acquire these abilities through training remain limited, particularly from a theoretical standpoint. This work investigates how transformers learn to solve symbolic multi-step reasoning problems through chain-of-thought processes, focusing on path-finding in trees. We analyze two intertwined tasks: a backward reasoning task, where the model outputs a path from a goal node to the root, and a more complex forward reasoning task, where the model implements two-stage reasoning by first identifying the goal-to-root path and then reversing it to produce the root-to-goal path. Our theoretical analysis, grounded in the dynamics of gradient descent, shows that trained one-layer transformers can provably solve both tasks with generalization guarantees to unseen trees. In particular, our multi-phase training dynamics for forward reasoning elucidate how different attention heads learn to specialize and coordinate autonomously to solve the two subtasks in a single autoregressive path. These results provide a mechanistic explanation of how trained transformers can implement sequential algorithmic procedures. Moreover, they offer insights into the emergence of reasoning abilities, suggesting that when tasks are structured to take intermediate chain-of-thought steps, even shallow multi-head transformers can effectively solve problems that would otherwise require deeper architectures.

PDF Details

TMLR Journal 2025 Journal Article

Pre-Training Representations of Binary Code Using Contrastive Learning

Yifan Zhang
Chen Huang
Yueke Zhang
Huajie Shao
Kevin Leach
Yu Huang

Binary code analysis and comprehension is critical to applications in reverse engineering and computer security tasks where source code is not available. Unfortunately, unlike source code, binary code lacks semantics and is more difficult for human engineers to understand and analyze. In this paper, we present ContraBin, a contrastive learning technique that integrates source code and comment information along with binaries to create an embedding capable of aiding binary analysis and comprehension tasks. Specifically, we present three components in ContraBin: (1) a primary contrastive learning method for initial pre-training, (2) a simplex interpolation method to integrate source code, comments, and binary code, and (3) an intermediate representation learning algorithm to train a binary code embedding. We further analyze the impact of human-written and synthetic comments on binary code comprehension tasks, revealing a significant performance disparity. While synthetic comments provide substantial benefits, human-written comments are found to introduce noise, even resulting in performance drops compared to using no comments. These findings reshape the narrative around the role of comment types in binary code analysis. We evaluate the effectiveness of ContraBin through four indicative downstream tasks related to binary code: algorithmic functionality classification, function name recovery, code summarization, and reverse engineering. The results show that ContraBin considerably improves performance on all four tasks, measured by accuracy, mean of average precision, and BLEU scores as appropriate. ContraBin is the first language representation model to incorporate source code, binary code, and comments into contrastive code representation learning and is intended to contribute to the field of binary code analysis. The dataset used in this study is available for further research.

PDF Details

JBHI Journal 2025 Journal Article

Predicting Mutation-Disease Associations Through Protein Interactions Via Deep Learning

Xue Li
Ben Cao
Jianmin Wang
Xiangyu Meng
Shuang Wang
Yu Huang
Enrico Petretto
Tao Song

Disease is one of the primary factors affecting life activities, with complex etiologies often influenced by gene expression and mutation. Currently, wet lab experiments have analyzed the mechanisms of mutations, but these are usually limited by the costs of wet experiments and constraints in sample types and scales. Therefore, this paper constructs a real-world mutation-induced disease dataset and proposes Capsule and Graph topology networks with Multi-head attention (CGM) to predict the mutation-disease associations. CGM can accurately predict protein mutation-disease associations, and to further elucidate the pathogenicity of protein mutations, we also verified that protein mutations lead to protein structural alterations by the model, which suggests that mutation-induced conformational changes may be an important pathogenic factor. Limited by the size of the mutated protein dataset, we also performed experiments on benchmark and imbalanced datasets, where CGM mined 22 unknown protein interaction pairs from the benchmark dataset, better illustrating the potential of CGM in predicting mutation-disease associations. In summary, this paper curates a real dataset. It proposes that CGM predicts protein mutations and disease associations, providing a novel tool for further understanding of biomolecular pathways and disease mechanisms.

Details DOI

AAAI Conference 2025 Conference Paper

STAMPsy: Towards SpatioTemporal-Aware Mixed-Type Dialogues for Psychological Counseling

Jieyi Wang
Yue Huang
Zeming Liu
Dexuan Xu
Chuan Wang
Xiaoming Shi
Ruiyuan Guan
Hongxing Wang

Online psychological counseling dialogue systems are trending, offering a convenient and accessible alternative to traditional in-person therapy. However, existing psychological counseling dialogue systems mainly focus on basic empathetic dialogue or QA with minimal professional knowledge and without goal guidance. In many real-world counseling scenarios, clients often seek multi-type help, such as diagnosis, consultation, therapy, console, and common questions, but existing dialogue systems struggle to combine different dialogue types naturally. In this paper, we identify this challenge as how to construct mixed-type dialogue systems for psychological counseling that enable clients to clarify their goals before proceeding with counseling. To mitigate the challenge, we collect a mixed-type counseling dialogues corpus termed STAMPsy, covering five dialogue types, task-oriented dialogue for diagnosis, knowledge-grounded dialogue, conversational recommendation, empathetic dialogue, and question answering, over 5,000 conversations. Moreover, spatiotemporal-aware knowledge enables systems to have world awareness and has been proven to affect one's mental health. Therefore, we link dialogues in STAMPsy to spatiotemporal state and propose a spatiotemporal-aware mixed-type psychological counseling dataset. Additionally, we build baselines on STAMPsy and develop an iterative self-feedback psychological dialogue generation framework, named Self-STAMPsy. Results indicate that clarifying dialogue goals in advance and utilizing spatiotemporal states are effective.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

Transformers Provably Learn Chain-of-Thought Reasoning with Length Generalization

Yu Huang
Zixin Wen
Aarti Singh
Yuejie Chi
Yuxin Chen

The ability to reason lies at the core of artificial intelligence (AI), and challenging problems usually call for deeper and longer reasoning to tackle. A crucial question about AI reasoning is whether models can extrapolate learned reasoning patterns to solve harder tasks with a longer chain-of-thought (CoT). In this work, we present a theoretical analysis of transformers learning on synthetic state-tracking tasks with gradient descent. Specifically: 1). We prove how the *algebraic structure* of state-tracking problems governs the length generalization of learned reasoning in transformers. In doing so, we formulate the **attention concentration** mechanism, linking the retrieval robustness of the attention layer to the task structure of long-context state tracking problems. 2). Moreover, we prove that a transformer can provably *self-improve* via a *recursive self-training* scheme that progressively extends the range of solvable problem lengths. We show that the model can achieve abilities outside the coverage of the base model in recursive training, different from prior theoretical works on self-improvement. To our knowledge, we provide the first *optimization guarantee* that constant-depth transformers provably learn $\text{NC}^1$-complete problems with CoT, significantly going beyond prior art confined in $\text{TC}^0$, unless the widely held conjecture $\text{TC}^0 \neq \text{NC}^1$ fails. Finally, we present a broad set of experiments supporting our theoretical results, confirming the length generalization behaviors and the mechanism of attention concentration.

PDF Details

AAAI Conference 2024 Conference Paper

A Learnable Discrete-Prior Fusion Autoencoder with Contrastive Learning for Tabular Data Synthesis

Rongchao Zhang
Yiwei Lou
Dexuan Xu
Yongzhi Cao
Hanpin Wang
Yu Huang

The actual collection of tabular data for sharing involves confidentiality and privacy constraints, leaving the potential risks of machine learning for interventional data analysis unsafely averted. Synthetic data has emerged recently as a privacy-protecting solution to address this challenge. However, existing approaches regard discrete and continuous modal features as separate entities, thus falling short in properly capturing their inherent correlations. In this paper, we propose a novel contrastive learning guided Gaussian Transformer autoencoder, termed GTCoder, to synthesize photo-realistic multimodal tabular data for scientific research. Our approach introduces a transformer-based fusion module that seamlessly integrates multimodal features, permitting for mining more informative latent representations. The attention within the fusion module directs the integrated output features to focus on critical components that facilitate the task of generating latent embeddings. Moreover, we formulate a contrastive learning strategy to implicitly constrain the embeddings from discrete features in the latent feature space by encouraging the similar discrete feature distributions closer while pushing the dissimilar further away, in order to better enhance the representation of the latent embedding. Experimental results indicate that GTCoder is effective to generate photo-realistic synthetic data, with interactive interpretation of latent embedding, and performs favorably against some baselines on most real-world and simulated datasets.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

In-Context Learning with Representations: Contextual Generalization of Trained Transformers

Tong Yang
Yu Huang
Yingbin Liang
Yuejie Chi

In-context learning (ICL) refers to a remarkable capability of pretrained large language models, which can learn a new task given a few examples during inference. However, theoretical understanding of ICL is largely under-explored, particularly whether transformers can be trained to generalize to unseen examples in a prompt, which will require the model to acquire contextual knowledge of the prompt for generalization. This paper investigates the training dynamics of transformers by gradient descent through the lens of non-linear regression tasks. The contextual generalization here can be attained via learning the template function for each task in-context, where all template functions lie in a linear space with $m$ basis functions. We analyze the training dynamics of one-layer multi-head transformers to {in-contextly} predict unlabeled inputs given partially labeled prompts, where the labels contain Gaussian noise and the number of examples in each prompt are not sufficient to determine the template. Under mild assumptions, we show that the training loss for a one-layer multi-head transformer converges linearly to a global minimum. Moreover, the transformer effectively learns to perform ridge regression over the basis functions. To our knowledge, this study is the first provable demonstration that transformers can learn contextual (i. e. , template) information to generalize to both unseen examples and tasks when prompts contain only a small number of query-answer pairs.

PDF Details DOI

NeurIPS Conference 2023 Conference Paper

ClimSim: A large multi-scale dataset for hybrid physics-ML climate emulation

Sungduk Yu
Walter Hannah
Liran Peng
Jerry Lin
Mohamed Aziz Bhouri
Ritwik Gupta
Björn Lütjens
Justus C. Will

Modern climate projections lack adequate spatial and temporal resolution due to computational constraints. A consequence is inaccurate and imprecise predictions of critical processes such as storms. Hybrid methods that combine physics with machine learning (ML) have introduced a new generation of higher fidelity climate simulators that can sidestep Moore's Law by outsourcing compute-hungry, short, high-resolution simulations to ML emulators. However, this hybrid ML-physics simulation approach requires domain-specific treatment and has been inaccessible to ML experts because of lack of training data and relevant, easy-to-use workflows. We present ClimSim, the largest-ever dataset designed for hybrid ML-physics research. It comprises multi-scale climate simulations, developed by a consortium of climate scientists and ML researchers. It consists of 5. 7 billion pairs of multivariate input and output vectors that isolate the influence of locally-nested, high-resolution, high-fidelity physics on a host climate simulator's macro-scale physical state. The dataset is global in coverage, spans multiple years at high sampling frequency, and is designed such that resulting emulators are compatible with downstream coupling into operational climate simulators. We implement a range of deterministic and stochastic regression baselines to highlight the ML challenges and their scoring. The data (https: //huggingface. co/datasets/LEAP/ClimSim_high-res) and code (https: //leap-stc. github. io/ClimSim) are released openly to support the development of hybrid ML-physics and high-fidelity climate simulations for the benefit of science and society.

PDF Details

TMLR Journal 2023 Journal Article

Online Min-max Problems with Non-convexity and Non-stationarity

Yu Huang
Yuan Cheng
Yingbin Liang
Longbo Huang

Online min-max optimization has recently gained considerable interest due to its rich applications to game theory, multi-agent reinforcement learning, online robust learning, etc. Theoretical understanding in this field has been mainly focused on convex-concave settings. Online min-max optimization with nonconvex geometries, which captures various online deep learning problems, has yet been studied so far. In this paper, we make the first effort and investigate online nonconvex-strongly-concave min-max optimization in the nonstationary environment. We first introduce a natural notion of local Nash equilibrium (NE)-regret, and then propose a novel algorithm coined TSODA to achieve the optimal regret. We further generalize our study to the setting with stochastic first-order feedback, and show that a variation of TSODA can also achieve the same optimal regret in expectation. Our theoretical results and the superior performance of the proposed method are further validated by empirical experiments. To our best knowledge, this is the first exploration of efficient online nonconvex min-max optimization.

PDF Details

JBHI Journal 2022 Journal Article

A Novel Constraint-Based Knee- Guided Neuroevolutionary Algorithm for Context-Specific ECG Early Classification

Yu Huang
Gary G. Yen
Vincent S. Tseng

Cardiovascular diseases (CVDs) are considered the greatest threat to human life according to World Health Organization. Early classification of CVDs and the appropriate follow-up treatment are crucial for preventing sudden deaths. Electrocardiogram (ECG) is one of the most common non-invasive tools used to evaluate the state of the heart, which can be exploited to automatically diagnose as well. However, the importance of diagnosing CVDs is varying in different context-specific scenarios. For example, ST-segment elevation (STE) is an acute myocardial infarction indicator for patients associated with chest pain and cardiac biomarker. In in-hospital healthcare, STE should be diagnosed with a higher priority than the other phenotypes of ECG. Hence, the context-specific requirements should be considered in ECG early classification problems. We formalize the ECG early classification problem as the context-specific time series classification problem. We propose a novel Constraint-based Knee-guided Neuroevolutionary Algorithm (CKNA) based on the Snippet Policy Networks V2 to solve this problem. To validate the proposed method, we perform a series of experiments on two public ECG datasets under various context-specific simulated scenarios after consulting with physicians specializing in the area. Experimental results show that CKNA significantly improves the average recall of disease classification by 5. 5% compared to the competing baseline under user-specified requirements. Moreover, experimental results prove that CKNA presents a feasible solution for the early classifying of cardiac arrhythmias under different user-specified scenarios.

Details DOI

NeurIPS Conference 2022 Conference Paper

Provable Generalization of Overparameterized Meta-learning Trained with SGD

Yu Huang
Yingbin Liang
Longbo Huang

Despite the empirical success of deep meta-learning, theoretical understanding of overparameterized meta-learning is still limited. This paper studies the generalization of a widely used meta-learning approach, Model-Agnostic Meta-Learning (MAML), which aims to find a good initialization for fast adaptation to new tasks. Under a mixed linear regression model, we analyze the generalization properties of MAML trained with SGD in the overparameterized regime. We provide both upper and lower bounds for the excess risk of MAML, which captures how SGD dynamics affect these generalization bounds. With such sharp characterizations, we further explore how various learning parameters impact the generalization capability of overparameterized MAML, including explicitly identifying typical data and task distributions that can achieve diminishing generalization error with overparameterization, and characterizing the impact of adaptation learning rate on both excess risk and the early stopping time. Our theoretical findings are further validated by experiments.

PDF Details

IJCAI Conference 2021 Conference Paper

GAEN: Graph Attention Evolving Networks

Min Shi
Yu Huang
Xingquan Zhu
Yufei Tang
Yuan Zhuang
Jianxun Liu

Real-world networked systems often show dynamic properties with continuously evolving network nodes and topology over time. When learning from dynamic networks, it is beneficial to correlate all temporal networks to fully capture the similarity/relevance between nodes. Recent work for dynamic network representation learning typically trains each single network independently and imposes relevance regularization on the network learning at different time steps. Such a snapshot scheme fails to leverage topology similarity between temporal networks for progressive training. In addition to the static node relationships within each network, nodes could show similar variation patterns (e. g. , change of local structures) within the temporal network sequence. Both static node structures and temporal variation patterns can be combined to better characterize node affinities for unified embedding learning. In this paper, we propose Graph Attention Evolving Networks (GAEN) for dynamic network embedding with preserved similarities between nodes derived from their temporal variation patterns. Instead of training graph attention weights for each network independently, we allow model weights to share and evolve across all temporal networks based on their respective topology discrepancies. Experiments and validations, on four real-world dynamic graphs, demonstrate that GAEN outperforms the state-of-the-art in both link prediction and node classification tasks.

PDF Details DOI

NeurIPS Conference 2021 Conference Paper

What Makes Multi-Modal Learning Better than Single (Provably)

Yu Huang
Chenzhuang Du
Zihui Xue
Xuanyao Chen
Hang Zhao
Longbo Huang

The world provides us with data of multiple modalities. Intuitively, models fusing data from different modalities outperform their uni-modal counterparts, since more information is aggregated. Recently, joining the success of deep learning, there is an influential line of work on deep multi-modal learning, which has remarkable empirical results on various applications. However, theoretical justifications in this field are notably lacking. Can multi-modal learning provably perform better than uni-modal? In this paper, we answer this question under a most popular multi-modal fusion framework, which firstly encodes features from different modalities into a common latent space and seamlessly maps the latent representations into the task space. We prove that learning with multiple modalities achieves a smaller population risk than only using its subset of modalities. The main intuition is that the former has a more accurate estimate of the latent space representation. To the best of our knowledge, this is the first theoretical treatment to capture important qualitative phenomena observed in real multi-modal applications from the generalization perspective. Combining with experiment results, we show that multi-modal learning does possess an appealing formal guarantee.

PDF Details

TIST Journal 2017 Journal Article

Daehr

Haoyi Xiong
Jinghe Zhang
Yu Huang
Kevin Leach
Laura E. Barnes

Electronic health records (EHR) provide a rich source of temporal data that present a unique opportunity to characterize disease patterns and risk of imminent disease. While many data-mining tools have been adopted for EHR-based disease early detection, linear discriminant analysis (LDA) is one of the most commonly used statistical methods. However, it is difficult to train an accurate LDA model for early disease diagnosis when too few patients are known to have the target disease. Furthermore, EHR data are heterogeneous with significant noise. In such cases, the covariance matrices used in LDA are usually singular and estimated with a large variance. This article presents Daehr, an extension of the LDA framework using electronic health record data to address these issues. Beyond existing LDA analyzers, we propose Daehr to (1) eliminate the data noise caused by the manual encoding of EHR data and (2) lower the variance of parameter (covariance matrices) estimation for LDA models when only a few patients’ EHR are available for training. To achieve these two goals, we designed an iterative algorithm to improve the covariance matrix estimation with embedded data-noise/parameter-variance reduction for LDA. We evaluated Daehr extensively using the College Health Surveillance Network, a large, real-world EHR dataset. Specifically, our experiments compared the performance of LDA to three baselines (i.e., LDA and its derivatives) in identifying college students at high risk for mental health disorders from 23 U.S. universities. Experimental results demonstrate Daehr significantly outperforms the three baselines by achieving 1.4%--19.4% higher accuracy and a 7.5%--43.5% higher F1-score.

Details DOI

YNIMG Journal 2016 Journal Article

The New York Head—A precise standardized volume conductor model for EEG source localization and tES targeting

Yu Huang
Lucas C. Parra
Stefan Haufe

In source localization of electroencephalograpic (EEG) signals, as well as in targeted transcranial electric current stimulation (tES), a volume conductor model is required to describe the flow of electric currents in the head. Boundary element models (BEM) can be readily computed to represent major tissue compartments, but cannot encode detailed anatomical information within compartments. Finite element models (FEM) can capture more tissue types and intricate anatomical structures, but with the higher precision also comes the need for semi-automated segmentation, and a higher computational cost. In either case, adjusting to the individual human anatomy requires costly magnetic resonance imaging (MRI), and thus head modeling is often based on the anatomy of an ‘arbitrary’ individual (e. g. Colin27). Additionally, existing reference models for the human head often do not include the cerebro-spinal fluid (CSF), and their field of view excludes portions of the head and neck—two factors that demonstrably affect current-flow patterns. Here we present a highly detailed FEM, which we call ICBM-NY, or "New York Head". It is based on the ICBM152 anatomical template (a non-linear average of the MRI of 152 adult human brains) defined in MNI coordinates, for which we extended the field of view to the neck and performed a detailed segmentation of six tissue types (scalp, skull, CSF, gray matter, white matter, air cavities) at 0. 5mm 3 resolution. The model was solved for 231 electrode locations. To evaluate its performance, additional FEMs and BEMs were constructed for four individual subjects. Each of the four individual FEMs (regarded as the ‘ground truth’) is compared to its BEM counterpart, the ICBM-NY, a BEM of the ICBM anatomy, an ‘individualized’ BEM of the ICBM anatomy warped to the individual head surface, and FEMs of the other individuals. Performance is measured in terms of EEG source localization and tES targeting errors. Results show that the ICBM-NY outperforms FEMs of mismatched individual anatomies as well as the BEM of the ICBM anatomy according to both criteria. We therefore propose the New York Head as a new standard head model to be used in future EEG and tES studies whenever an individual MRI is not available. We release all model data online at neuralengr. com/nyhead/ to facilitate broad adoption.

Details DOI

YNIMG Journal 2013 Journal Article

Targeted transcranial direct current stimulation for rehabilitation after stroke

Jacek P. Dmochowski
Abhishek Datta
Yu Huang
Jessica D. Richardson
Marom Bikson
Julius Fridriksson
Lucas C. Parra

Transcranial direct current stimulation (tDCS) is being investigated as an adjunctive technique to behavioral rehabilitation treatment after stroke. The conventional “dosage”, consisting of a large (25cm2) anode over the target with the cathode over the contralateral hemisphere, has been previously shown to yield broadly distributed electric fields whose intensities at the target region are less than maximal. Here, we report the results of a systematic targeting procedure with small “high-definition” electrodes that was used in preparation for a pilot study on 8 stroke patients with chronic aphasia. We employ functional and anatomical magnetic resonance imagery (fMRI/MRI) to define a target and optimize (with respect to the electric field magnitude at the target) the electrode configuration, respectively, and demonstrate that electric field strengths in targeted cortex can be substantially increased (63%) over the conventional approach. The optimal montage exhibits significant variation across subjects as well as when perturbing the target location within a subject. However, for each displacement of the target co-ordinates, the algorithm is able to determine a montage which delivers a consistent amount of current to that location. These results demonstrate that MRI-based models of current flow yield maximal stimulation of target structures, and as such, may aid in reliably assessing the efficacy of tDCS in neuro-rehabilitation.

Details DOI