Birdie Blog

Thinking will not overcome fear but action will.

DIFUSCO Graph-based Diffusion Solvers for Combinatorial Optimization

NIPS2023 扩散模型

DIFUSCO: Graph-based Diffusion Solvers for Combinatorial Optimization nips23 Carnegie Mellon University 开源：Edward-Sun/DIFUSCO: Code of NeurIPS paper: arxiv.org/abs/2302.08224 摘要 基于神经网络的组合优化（Co...

Posted by Birdie on November 7, 2025

Improving Generalization of Neural Combinatorial Optimization for Vehicle Routing Problems via Test-Time Projection Learning

NIPS2025 EoH驱动的映射算子做大规模

Improving Generalization of Neural Combinatorial Optimization for Vehicle Routing Problems via Test-Time Projection Learning 王振坤课题组 NIPS2025 代码：https://github.com/CIAM-Group/TTPL 摘要 神经组合优化（NCO...

Posted by Birdie on November 5, 2025

TODO Enhancing LLM Alignment with Ternary Preferences

ICLR2025 偏好优化 平局

偏好优化理论基础 BT模型 BT模型是一种用于表示实例、团队或对象之间成对比较结果的概率模型。它估计了排序关系 $i \succ j$ 为真的概率，其中符号 $\succ$ 表示偏好或排序关系，例如实例 $i$ 被偏好于 $j$。 BT 模型的计算如下所示，其中两个竞争者的正向强度分别表示为 $\lambda_1$ 和 $\lambda_2$，$r_{1,2}$ 表示第一个竞争者在比...

Posted by Birdie on May 30, 2025

GLOP Learning Global Partition and Local Construction for Solving Large-scale Routing Problems in Real-time

AAAI2024 通过分治将大规模/多任务变成多个开源TSP

GLOP: Learning Global Partition and Local Construction for Solving Large-scale Routing Problems in Real-time AAAI2024 开源：https://github.com/henry-yeh/GLOP 摘要 最近的端到端神经求解器在处理小型路线规划问题上展现出了潜力，但在实...

Posted by Birdie on April 30, 2025

PolyNet Learning Diverse Solution Strategies for Neural Combinatorial Optimization

ICLR2025 扩大搜索空间

PolyNet: Learning Diverse Solution Strategies for Neural Combinatorial Optimization ICLR2025 Bielefeld University，Germany 摘要 基于强化学习的方法在构建组合优化问题的解决方案方面，正迅速接近人工设计算法的性能水平。为了进一步缩小这一差距，基于学习的方法必须在搜索过...

Posted by Birdie on April 29, 2025

Rl4co an extensive reinforcement learning for combinatorial optimization benchmark

ICLR2024 通用、全面的RL4CO库

Rl4co: an extensive reinforcement learning for combinatorial optimization benchmark 一个用RL解决CO问题的算法库 开源：ai4co/rl4co: A PyTorch library for all things Reinforcement Learning (RL) for Combinatoria...

Posted by Birdie on April 28, 2025

INViT A Generalizable Routing Problem Solver with Invariant Nested View Transformer

ICML2024 大规模+多分布

INViT: A Generalizable Routing Problem Solver with Invariant Nested View Transformer 上海交通大学密歇根联合研究所，昆山杜克大学 ICML2024 开源：Kasumigaoka-Utaha/INViT: Official Implementation of the paper: INViT: A Gen...

Posted by Birdie on April 28, 2025

Distilling Autoregressive Models to Obtain High-Performance Non-Autoregressive Solvers for Vehicle Routing Problems with Faster Inference Speed

AAAI2024 引导知识蒸馏 非自回归模型

Distilling Autoregressive Models to Obtain High-Performance Non-Autoregressive Solvers for Vehicle Routing Problems with Faster Inference Speed AAAI2024 省流 就看方法论里面的那张图，做了一个从自回归引导到非自回归的知识蒸馏。 摘...

Posted by Birdie on April 28, 2025

Learning What to Defer for Maximum Independent Sets

ICML2020 延迟决策，鼓励解多样性

Learning What to Defer for Maximum Independent Sets ICML2020 KAIST 摘要 设计高效的组合优化算法在各个科学领域中普遍存在。近年来，深度强化学习（DRL）框架作为一种新方法受到了相当多的关注：它们可以在依赖较少目标问题的专业知识的情况下自动化求解器的设计。然而，现有的DRL求解器通过与解中元素数量成比例的阶段数来确定解，...

Posted by Birdie on April 27, 2025

Memory-Enhanced Neural Solvers for Efficient Adaptation in Combinatorial Optimization

Arxiv2024.10 经验池

Memory-Enhanced Neural Solvers for Efficient Adaptation in Combinatorial Optimization Arxiv 2024.10.7 InstaDeep 开源：instadeepai/memento: Official Implementation of Memento 摘要 组合优化对许多现实世界的应用至关重要...

Posted by Birdie on April 25, 2025

FEATURED TAGS
神经网络 机器学习 机器人 webot 通信原理 并行计算 信息安全 自然语言处理 数据库 数字图像处理 软件工程 编译原理 强化学习 数字电路 proteus 计算机组成原理 VIVADO JAVA 论文阅读 组合优化 NIPS AAAI ICML IJCAI 隐私保护 ICLR
ABOUT ME

Birdie-Go

✉️ fangzhh6@mail2.sysu.edu.cn