Author name cluster

Bei Yu

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.

18 papers

1 author row

AAAI Conference 2026 Conference Paper

KCLNet: Electrically Equivalence-Oriented Graph Representation Learning for Analog Circuits

Peng Xu
Yapeng Li
Tinghuan Chen
Tsung-Yi Ho
Bei Yu

Digital circuit representation learning has made remarkable progress in electronic design automation, effectively supporting critical tasks such as testability analysis and logic reasoning. However, representation learning for analog circuits remains challenging due to their continuous electrical characteristics compared to the discrete states of digital circuits. This paper presents a direct current (DC) electrically equivalent-oriented analog representation learning framework, named KCLNet. We will open-source the dataset and code upon publication. It comprises an asynchronous graph neural network structure with electrically-simulated message passing and a representation learning method inspired by Kirchhoff's Current Law (KCL). This method maintains the orderliness of the circuit embedding space by enforcing the equality of the sum of outgoing and incoming current embeddings at each node, which significantly enhances the generalization ability of circuit embeddings. KCLNet offers a novel and effective solution for analog circuit representation learning with electrical constraints preserved. Experimental results demonstrate that our method achieves significant performance in a variety of downstream tasks, e.g., analog circuit classification, subcircuit detection, and circuit edit distance prediction.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

DLoFT: Gradient-Decoupled Fine-Tuning for Generalizable Long Chain-of-Thought Reasoning

Sitong Wu
Haoru Tan
Jingyao Li
Shaofeng Zhang
Xiaojuan Qi
Bei Yu
Jiaya Jia

Long chain-of-thought (LongCoT) has emerged as a powerful reasoning paradigm for enabling large language models (LLMs) to solve complex tasks through a systematic and thorough thinking phase. Although supervised fine-tuning (SFT) on high-quality LongCoT traces has proven effective to activate LongCoT abilities, we find that models trained in this way tend to overfit problem-specific knowledge and heuristics, leading to degraded out-of-distribution performance. To address this issue, we propose a Decoupled LongCoT Fine-Tuning (DLoFT) algorithm, which enables the model to learn generalizable LongCoT reasoning abilities while preventing overfitting to the reasoning content with problem-specific information. The key idea is to decouple the gradient into two orthogonal components: 1) a paradigm-relevant gradient corresponding to the general LongCoT paradigm and 2) a content-relevant gradient reflecting the problem-specific information, where only the former gradient is used to update model parameters. Specifically, by leveraging the unique two-phase composition (thinking and solution) of the LongCoT response, our gradient decoupling mechanism isolates the content-relevant gradient via a projection operation and separates the paradigm-relevant gradient through orthogonalization. Our DLoFT ensures the model concentrate on internalizing the LongCoT paradigm rather than memorizing problem-specific knowledge and heuristics. Extensive experiments demonstrate that our DLoFT significantly improves the generalization behavior of LongCoT abilities compared to SFT while maintaining strong in-distribution performance.

PDF Details

AAAI Conference 2025 Conference Paper

FlexPose: Pose Distribution Adaptation with Limited Guidance

Zixiao Wang
Junwu Weng
Mengyuan Liu
Bei Yu

Numerous well-annotated human key-point datasets are publicly available to date. However, annotating human poses for newly collected images is still a costly and time-consuming progress. Pose distributions from different datasets share similar pose hinge-structure priors with different geometric transformations, such as pivot orientation, joint rotation, and bone length ratio. The difference between Pose distributions is essentially the difference between the transformation distributions. Inspired by this fact, we propose a method to calibrate a pre-trained pose generator in which the pose prior has already been learned to an adapted one following a new pose distribution. We treat the representation of human pose joint coordinates as skeleton image and transfer a pre-trained pose annotation generator with only a few annotation guidance. By fine-tuning a limited number of linear layers that closely related to the pose transformation, the adapted generator is able to produce any number of pose annotations that are similar to the target poses. We evaluate our proposed method, FlexPose, on several cross-dataset settings both qualitatively and quantitatively, which demonstrates that our approach achieves state-of-the-art performance compared to the existing generative-model-based transfer learning methods when given limited annotation guidance.

PDF Details DOI

AAAI Conference 2025 Conference Paper

Intelligent OPC Engineer Assistant for Semiconductor Manufacturing

Guojin Chen
Haoyu Yang
Bei Yu
Haoxing Ren

Advancements in chip design and manufacturing have enabled the processing of complex tasks such as deep learning and natural language processing, paving the way for the development of artificial general intelligence (AGI). AI, on the other hand, can be leveraged to innovate and streamline semiconductor technology from planning and implementation to manufacturing. In this paper, we present Intelligent OPC Engineer Assistant, an AI/LLM-powered methodology designed to solve the core manufacturing-aware optimization problem known as Optical Proximity Correction (OPC). The methodology involves a reinforcement learning-based OPC recipe search and a customized multi-modal agent system for recipe summarization. Experiments demonstrate that our methodology can efficiently build OPC recipes on various chip designs with specially handled design topologies, a task that typically requires the full-time effort of OPC engineers with years of experience.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

LithoSim: A Large, Holistic Lithography Simulation Benchmark for AI-Driven Semiconductor Manufacturing

Hongquan He
Zhen Wang
Jingya Wang
Tao Wu
Xuming He
Bei Yu
Jingyi Yu
Hao GENG

Lithography orchestrates a symphony of light, mask and photochemicals to transfer the integrated circuit patterns onto the wafer. Lithography simulation serves as the critical nexus between circuit design and manufacturing, where its speed and accuracy fundamentally govern the optimization quality of downstream resolution enhancement techniques (RET). While machine learning promises to circumvent computational limitations of lithography process through data-driven or physics-informed approximations of computational lithography, existing simulators suffer from inadequate lithographic awareness due to insufficient training data capturing essential process variations and mask correction rules. We present LithoSim, the most comprehensive lithography simulation benchmark to date, featuring over $4$ million high-resolution input-output pairs with rigorous physical correspondence. The dataset systematically incorporates alterable optical source distributions, metal and via mask topologies with optical proximity correction (OPC) variants, and process windows reflecting fab-realistic variations. By integrating domain-specific metrics spanning AI performance and lithographic fidelity, LithoSim establishes a unified evaluation framework for data-driven and physics-informed computational lithography. The data (https: //huggingface. co/datasets/grandiflorum/LithoSim), code (https: //dw-hongquan. github. io/LithoSim), and pre-trained models (https: //huggingface. co/grandiflorum/LithoSim) are released openly to support the development of hybrid ML-based and high-fidelity lithography simulation for the benefit of semiconductor manufacturing.

PDF Details

IJCAI Conference 2025 Conference Paper

Low-Light Video Enhancement via Spatial-Temporal Consistent Decomposition

Xiaogang Xu
Kun Zhou
Tao Hu
Jiafei Wu
Ruixing Wang
Hao Peng
Bei Yu

Low-Light Video Enhancement (LLVE) seeks to restore dynamic or static scenes plagued by severe invisibility and noise. In this paper, we present an innovative video decomposition strategy that incorporates view-independent and view-dependent components to enhance the performance of LLVE. We leverage dynamic cross-frame correspondences for the view-independent term (which primarily captures intrinsic appearance) and impose a scene-level continuity constraint on the view-dependent term (which mainly describes the shading condition) to achieve consistent and satisfactory decomposition results. To further ensure consistent decomposition, we introduce a dual-structure enhancement network featuring a cross-frame interaction mechanism. By supervising different frames simultaneously, this network encourages them to exhibit matching decomposition features. This mechanism can seamlessly integrate with encoder-decoder single-frame networks, incurring minimal additional parameter costs. Extensive experiments are conducted on widely recognized LLVE benchmarks, covering diverse scenarios. Our framework consistently outperforms existing methods, establishing a new SOTA performance.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

On-Policy Optimization with Group Equivalent Preference for Multi-Programming Language Understanding

Haoyuan Wu
Rui Ming
Jilong Gao
Hangyu Zhao
Xueyi Chen
Yikai Yang
Haisheng Zheng
Zhuolun He

Large language models (LLMs) achieve remarkable performance in code generation tasks. However, a significant performance disparity persists between popular programming languages (e. g. , Python, C++) and others. To address this capability gap, we leverage the code translation task to train LLMs, thereby facilitating the transfer of coding proficiency across diverse programming languages. Moreover, we introduce OORL for training, a novel reinforcement learning (RL) framework that integrates on-policy and off-policy strategies. Within OORL, on-policy RL is applied during code translation, guided by a rule-based reward signal derived from unit tests. Complementing this coarse-grained rule-based reward, we propose Group Equivalent Preference Optimization (GEPO), a novel preference optimization method. Specifically, GEPO trains the LLM using intermediate representations (IRs) groups. LLMs can be guided to discern IRs equivalent to the source code from inequivalent ones, while also utilizing signals about the mutual equivalence between IRs within the group. This process allows LLMs to capture nuanced aspects of code functionality. By employing OORL for training with code translation tasks, LLMs improve their recognition of code functionality and their understanding of the relationships between code implemented in different languages. Extensive experiments demonstrate that our OORL for LLMs training with code translation tasks achieves significant performance improvements on code benchmarks across multiple programming languages.

PDF Details

NeurIPS Conference 2025 Conference Paper

PermLLM: Learnable Channel Permutation for N:M Sparse Large Language Models

Lancheng Zou
Shuo Yin
Zehua Pei
Tsung-Yi Ho
Farzan Farnia
Bei Yu

Channel permutation is a powerful technique for enhancing the accuracy of N: M sparse models by reordering the channels of weight matrices to prioritize the retention of important weights. However, traditional channel permutation methods rely on handcrafted quality metrics, which often fail to accurately capture the true impact of pruning on model performance. To address this limitation, we propose PermLLM, a novel post-training pruning framework that introduces learnable channel permutation (LCP) for N: M sparsity. LCP leverages Sinkhorn normalization to transform discrete permutation matrices into differentiable soft permutation matrices, enabling end-to-end optimization. Additionally, PermLLM incorporates an efficient block-wise channel permutation strategy, which significantly reduces the number of learnable parameters and computational complexity. PermLLM seamlessly integrates with existing one-shot pruning methods to adaptively optimize channel permutations, effectively mitigating pruning-induced errors. Extensive experiments on the LLaMA series, Qwen, and OPT models demonstrate that PermLLM achieves superior performance in optimizing N: M sparse models.

PDF Details

NeurIPS Conference 2025 Conference Paper

VisionThink: Smart and Efficient Vision Language Model via Reinforcement Learning

Senqiao Yang
Junyi Li
Xin Lai
Jinming Wu
Wei Li
Zejun Ma
Bei Yu
Hengshuang Zhao

Recent advancements in vision-language models (VLMs) have improved performance by increasing the number of visual tokens, which are often significantly longer than text tokens. However, we observe that most real-world scenarios do not require such an extensive number of visual tokens. While the performance drops significantly in a small subset of OCR-related tasks, models still perform accurately in most other general VQA tasks with only 1/4 resolution. Therefore, we propose to dynamically process distinct samples with different resolutions, and present a new paradigm for visual token compression, namely, VisionThink. It starts with a downsampled image and smartly decides whether it is sufficient for problem solving. Otherwise, the model could output a special token to request the higher-resolution image. Compared to existing Efficient VLM methods that compress tokens using fixed pruning ratios or thresholds, VisionThink autonomously decides whether to compress tokens case by case. As a result, it demonstrates strong fine-grained visual understanding capability on OCR-related tasks, and meanwhile saves substantial visual tokens on simpler tasks. We adopt reinforcement learning and propose the LLM-as-Judge strategy to successfully apply RL to general VQA tasks. Moreoever, we carefully design a reward function and penalty mechanism to achieve a stable and reasonable image resize call ratio. Extensive experiments demonstrate the superiority, efficiency, and effectiveness of our method. All our code and data are open-sourced.

PDF Details

NeurIPS Conference 2024 Conference Paper

Decoupled Kullback-Leibler Divergence Loss

Jiequan Cui
Zhuotao Tian
Zhisheng Zhong
Xiaojuan Qi
Bei Yu
Hanwang Zhang

In this paper, we delve deeper into the Kullback–Leibler (KL) Divergence loss and mathematically prove that it is equivalent to the Decoupled Kullback-Leibler (DKL) Divergence loss that consists of 1) a weighted Mean Square Error ($\mathbf{w}$MSE) loss and 2) a Cross-Entropy loss incorporating soft labels. Thanks to the decomposed formulation of DKL loss, we have identified two areas for improvement. Firstly, we address the limitation of KL/DKL in scenarios like knowledge distillation by breaking its asymmetric optimization property. This modification ensures that the $\mathbf{w}$MSE component is always effective during training, providing extra constructive cues. Secondly, we introduce class-wise global information into KL/DKL to mitigate bias from individual samples. With these two enhancements, we derive the Improved Kullback–Leibler (IKL) Divergence loss and evaluate its effectiveness by conducting experiments on CIFAR-10/100 and ImageNet datasets, focusing on adversarial training, and knowledge distillation tasks. The proposed approach achieves new state-of-the-art adversarial robustness on the public leaderboard --- \textit{RobustBench} and competitive performance on knowledge distillation, demonstrating the substantial practical merits. Our code is available at https: //github. com/jiequancui/DKL.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

Improving Neural ODE Training with Temporal Adaptive Batch Normalization

Su Zheng
Zhengqi Gao
Fan-Keng Sun
Duane S. Boning
Bei Yu
Martin Wong

Neural ordinary differential equations (Neural ODEs) is a family of continuous-depth neural networks where the evolution of hidden states is governed by learnable temporal derivatives. We identify a significant limitation in applying traditional Batch Normalization (BN) to Neural ODEs, due to a fundamental mismatch --- BN was initially designed for discrete neural networks with no temporal dimension, whereas Neural ODEs operate continuously over time. To bridge this gap, we introduce temporal adaptive Batch Normalization (TA-BN), a novel technique that acts as the continuous-time analog to traditional BN. Our empirical findings reveal that TA-BN enables the stacking of more layers within Neural ODEs, enhancing their performance. Moreover, when confined to a model architecture consisting of a single Neural ODE followed by a linear layer, TA-BN achieves 91. 1\% test accuracy on CIFAR-10 with 2. 2 million parameters, making it the first \texttt{unmixed} Neural ODE architecture to approach MobileNetV2-level parameter efficiency. Extensive numerical experiments on image classification and physical system modeling substantiate the superiority of TA-BN compared to baseline methods.

PDF Details DOI

AAAI Conference 2024 Conference Paper

p-Laplacian Adaptation for Generative Pre-trained Vision-Language Models

Haoyuan Wu
Xinyun Zhang
Peng Xu
Peiyu Liao
Xufeng Yao
Bei Yu

Vision-Language models (VLMs) pre-trained on large corpora have demonstrated notable success across a range of downstream tasks. In light of the rapidly increasing size of pre-trained VLMs, parameter-efficient transfer learning (PETL) has garnered attention as a viable alternative to full fine-tuning. One such approach is the adapter, which introduces a few trainable parameters into the pre-trained models while preserving the original parameters during adaptation. In this paper, we present a novel modeling framework that recasts adapter tuning after attention as a graph message passing process on attention graphs, where the projected query and value features and attention matrix constitute the node features and the graph adjacency matrix, respectively. Within this framework, tuning adapters in VLMs necessitates handling heterophilic graphs, owing to the disparity between the projected query and value space. To address this challenge, we propose a new adapter architecture, p-adapter, which employs p-Laplacian message passing in Graph Neural Networks (GNNs). Specifically, the attention weights are re-normalized based on the features, and the features are then aggregated using the calibrated attention matrix, enabling the dynamic exploitation of information with varying frequencies in the heterophilic attention graphs. We conduct extensive experiments on different pre-trained VLMs and multi-modal tasks, including visual question answering, visual entailment, and image captioning. The experimental results validate our method's significant superiority over other PETL methods. Our code is available at https://github.com/wuhy68/p-Adapter/.

PDF Details DOI

AAAI Conference 2024 Conference Paper

Progressively Knowledge Distillation via Re-parameterizing Diffusion Reverse Process

Xufeng Yao
Fanbin Lu
Yuechen Zhang
Xinyun Zhang
Wenqian Zhao
Bei Yu

Knowledge distillation aims at transferring knowledge from the teacher model to the student one by aligning their distributions. Feature-level distillation often uses L2 distance or its variants as the loss function, based on the assumption that outputs follow normal distributions. This poses a significant challenge when distribution gaps are substantial since this loss function ignores the variance term. To address the problem, we propose to decompose the transfer objective into small parts and optimize it progressively. This process is inspired by diffusion models from which the noise distribution is mapped to the target distribution step by step. However, directly employing diffusion models is impractical in the distillation scenario due to its heavy reverse process. To overcome this challenge, we adopt the structural re-parameterization technique to generate multiple student features to approximate the teacher features sequentially. The multiple student features are combined linearly in inference time without extra cost. We present extensive experiments performed on various transfer scenarios, such as CNN-to-CNN and Transformer-to-CNN, that validate the effectiveness of our approach.

PDF Details DOI

AAAI Conference 2024 Conference Paper

Towards Automated RISC-V Microarchitecture Design with Reinforcement Learning

Chen Bai
Jianwang Zhai
Yuzhe Ma
Bei Yu
Martin D. F. Wong

Microarchitecture determines the implementation of a microprocessor. Designing a microarchitecture to achieve better performance, power, and area (PPA) trade-off has been increasingly difficult. Previous data-driven methodologies hold inappropriate assumptions and lack more tightly coupling with expert knowledge. This paper proposes a novel reinforcement learning-based (RL) solution that addresses these limitations. With the integration of microarchitecture scaling graph, PPA preference space embedding, and proposed lightweight environment in RL, experiments using commercial electronic design automation (EDA) tools show that our method achieves an average PPA trade-off improvement of 16.03% than previous state-of-the-art approaches with 4.07× higher efficiency. The solution qualities outperform human implementations by at most 2.03× in the PPA trade-off.

PDF Details DOI

AAAI Conference 2023 Conference Paper

AutoGraph: Optimizing DNN Computation Graph for Parallel GPU Kernel Execution

Yuxuan Zhao
Qi Sun
Zhuolun He
Yang Bai
Bei Yu

Deep learning frameworks optimize the computation graphs and intra-operator computations to boost the inference performance on GPUs, while inter-operator parallelism is usually ignored. In this paper, a unified framework, AutoGraph, is proposed to obtain highly optimized computation graphs in favor of parallel executions of GPU kernels. A novel dynamic programming algorithm, combined with backtracking search, is adopted to explore the optimal graph optimization solution, with the fast performance estimation from the mixed critical path cost. Accurate runtime information based on GPU Multi-Stream launched with CUDA Graph is utilized to determine the convergence of the optimization. Experimental results demonstrate that our method achieves up to 3.47x speedup over existing graph optimization methods. Moreover, AutoGraph outperforms state-of-the-art parallel kernel launch frameworks by up to 1.26x.

PDF Details DOI

IJCAI Conference 2023 Conference Paper

FastGR: Global Routing on CPU-GPU with Heterogeneous Task Graph Scheduler (Extended Abstract)

Siting Liu
Yuan Pu
Peiyu Liao
Hongzhong Wu
Rui Zhang
Zhitang Chen
Wenlong Lv
Yibo Lin

Running time is a key metric across the standard physical design flow stages. However, with the rapid growth in design sizes, routing runtime has become the runtime bottleneck in the physical design flow. To improve the effectiveness of the modern global router, we propose a global routing framework with GPU-accelerated routing algorithms and a heterogeneous task graph scheduler, called FastGR. Its runtime-oriented version FastGRL achieves 2. 489× speedup compared with the state-of-the-art global router. Furthermore, the GPU-accelerated L-shape pattern routing used in FastGRL can contribute to 9. 324× speedup over the sequential algorithm on CPU. Its quality-oriented version FastGRH offers further quality improvement over FastGRL with similar acceleration.

PDF Details DOI

NeurIPS Conference 2023 Conference Paper

LithoBench: Benchmarking AI Computational Lithography for Semiconductor Manufacturing

Su Zheng
Haoyu Yang
Binwu Zhu
Bei Yu
Martin Wong

Computational lithography provides algorithmic and mathematical support for resolution enhancement in optical lithography, which is the critical step in semiconductor manufacturing. The time-consuming lithography simulation and mask optimization processes limit the practical application of inverse lithography technology (ILT), a promising solution to the challenges of advanced-node lithography. Although various machine learning methods for ILT have shown promise for reducing the computational burden, this field is in lack of a dataset that can train the models thoroughly and evaluate the performance comprehensively. To boost the development of AI-driven computational lithography, we present the LithoBench dataset, a collection of circuit layout tiles for deep-learning-based lithography simulation and mask optimization. LithoBench consists of more than 120k tiles that are cropped from real circuit designs or synthesized according to the layout topologies of famous ILT testcases. The ground truths are generated by a famous lithography model in academia and an advanced ILT method. Based on the data, we provide a framework to design and evaluate deep neural networks (DNNs) with the data. The framework is used to benchmark state-of-the-art models on lithography simulation and mask optimization. We hope LithoBench can promote the research and development of computational lithography. LithoBench is available at https: //anonymous. 4open. science/r/lithobench-APPL.

PDF Details

AAAI Conference 2022 Conference Paper

Context-Based Contrastive Learning for Scene Text Recognition

Xinyun Zhang
Binwu Zhu
Xufeng Yao
Qi Sun
Ruiyu Li
Bei Yu

Pursuing accurate and robust recognizers has been a longlasting goal for scene text recognition (STR) researchers. Recently, attention-based methods have demonstrated their effectiveness and achieved impressive results on public benchmarks. The attention mechanism enables models to recognize scene text with severe visual distortions by leveraging contextual information. However, recent studies revealed that the implicit over-reliance of context leads to catastrophic out-ofvocabulary performance. On the contrary to the superior accuracy of the seen text, models are prone to misrecognize unseen text even with good image quality. We propose a novel framework, Context-based contrastive learning (ConCLR), to alleviate this issue. Our proposed method first generates characters with different contexts via simple image concatenation operations and then optimizes contrastive loss on their embeddings. By pulling together clusters of identical characters within various contexts and pushing apart clusters of different characters in embedding space, ConCLR suppresses the side-effect of overfitting to specific contexts and learns a more robust representation. Experiments show that ConCLR significantly improves out-of-vocabulary generalization and achieves stateof-the-art performance on public benchmarks together with attention-based recognizers.

PDF Details