非凸优化论文趋势 2023–2026

主要结论

• 非凸优化论文从 2023 年的 346 篇增长到 2026 年的 757 篇，三年间扩大到约 2.2 倍。
• NeurIPS 是发非凸优化论文最多的会议，2023–2026 年共 958 篇，占四个会议总量的 41%。
• 交叉方向里涨得最快的是扩散模型：同时属于两个主题的论文从 2023 年的 2 篇增至 2026 年的 44 篇。
• 在「优化与缩放」的 5 个主题里，它的 2026 年论文数排第 2；同组增长最快的缩放定律是 4.1 倍，非凸优化是 2.2 倍。

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

• NeurIPS 2026$\gamma$-weakly $\theta$-up-concavity: A Unified Framework for Non-Convex Optimization Beyond DR-Submodular and OSS Functions
• ICML 2026$\texttt{MetaDistill}$: Unlocking the Performance Ceiling for Pretrained Optimizers
• ICLR 2026Behavioral Embeddings of Programs: A Quasi-Dynamic Approach for Optimization Prediction
• CVPR 2026Coverage Optimization for Camera View Selection
• NeurIPS 2025Beyond Random: Automatic Inner-loop Optimization in Dataset Distillation
• ICML 2025A Bregman Proximal Viewpoint on Neural Operators
• ICLR 2025A Truncated Newton Method for Optimal Transport
• CVPR 2025Balanced Rate-Distortion Optimization in Learned Image Compression
• NeurIPS 2024A survey and benchmark of high-dimensional Bayesian optimization of discrete sequences
• ICML 2024A Doubly Recursive Stochastic Compositional Gradient Descent Method for Federated Multi-Level Compositional Optimization
• ICLR 2024Bayesian Coreset Optimization for Personalized Federated Learning
• CVPR 2024FaceCom: Towards High-fidelity 3D Facial Shape Completion via Optimization and Inpainting Guidance
• NeurIPS 2023A Computation and Communication Efficient Method for Distributed Nonconvex Problems in the Partial Participation Setting
• ICML 2023A Distribution Optimization Framework for Confidence Bounds of Risk Measures
• ICLR 2023A view of mini-batch SGD via generating functions: conditions of convergence, phase transitions, benefit from negative momenta.
• CVPR 2023Attribute-Preserving Face Dataset Anonymization via Latent Code Optimization
• NeurIPS 2026A Doubly Smoothed Decentralized Stochastic Minimax Optimization Algorithm
• ICML 2026A Constrained Optimization Perspective of Unrolled Transformers
• ICLR 2026BEP: A Binary Error Propagation Algorithm for Binary Neural Networks Training
• CVPR 2026Fully Decentralized Certified Unlearning

全部非凸优化论文

共 757 篇，第 3 / 16 页

101. ICLR 2026FedMuon: Federated Learning with Bias-corrected LMO-based Optimization
102. ICLR 2026FFT-based Dynamic Subspace Selection for Low-Rank Adaptive Optimization of Large Language Models
103. ICLR 2026Fine-tuning Quantized Neural Networks with Zeroth-order Optimization
104. ICLR 2026From Fields to Random Trees
105. ICLR 2026From Sequential to Parallel: Reformulating Dynamic Programming as GPU Kernels for Large-Scale Stochastic Combinatorial Optimization
106. ICLR 2026From Sorting Algorithms to Scalable Kernels: Bayesian Optimization in High-Dimensional Permutation Spaces
107. ICLR 2026Gelato: Graph Edit Distance via Autoregressive Neural Combinatorial Optimization
108. ICLR 2026Generative Bayesian Optimization: Generative Models as Acquisition Functions
109. ICLR 2026Geometric-Mean Policy Optimization
110. ICLR 2026GIT-BO: High-Dimensional Bayesian Optimization with Tabular Foundation Models
111. ICLR 2026Global Resolution: Optimal Multi-Draft Speculative Sampling via Convex Optimization
112. ICLR 2026Globally aware optimization with resurgence
113. ICLR 2026GoalRank: Group-Relative Optimization for a Large Ranking Model
114. ICLR 2026Gradient Descent with Large Step Sizes: Chaos and Fractal Convergence Region
115. ICLR 2026Gradient-Based Diversity Optimization with Differentiable Top-$k$ Objective
116. ICLR 2026Gradient-Normalized Smoothness for Optimization with Approximate Hessians
117. ICLR 2026Group Verification-based Policy Optimization for Interactive Coding Agents
118. ICLR 2026Group-Normalized Implicit Value Optimization for Language Models
119. ICLR 2026Guided Policy Optimization under Partial Observability
120. ICLR 2026Harmonized Cone for Feasible and Non-conflict Directions in Training Physics-Informed Neural Networks
121. ICLR 2026High Probability Bounds for Non-Convex Stochastic Optimization with Momentum
122. ICLR 2026High-Probability Bounds for the Last Iterate of Clipped SGD
123. ICLR 2026Hinge Regression Tree: A Newton Method for Oblique Regression Tree Splitting
124. ICLR 2026How does the optimizer implicitly bias the model merging loss landscape?
125. ICLR 2026Implicit Bias and Loss of Plasticity in Matrix Completion: Depth Promotes Low-Rankness
126. ICLR 2026Improved $\ell_{p}$ Regression via Iteratively Reweighted Least Squares
127. ICLR 2026Improving Online-to-Nonconvex Conversion for Smooth Optimization via Double Optimism
128. ICLR 2026Incorporating Expert Priors into Bayesian Optimization via Dynamic Mean Decay
129. ICLR 2026Inpainting-Guided Policy Optimization for Diffusion Large Language Models
130. ICLR 2026Is On-Policy Data always the Best Choice for Direct Preference Optimization-Based LM Alignment?
131. ICLR 2026Joint Distribution–Informed Shapley Values for Sparse Counterfactual Explanations
132. ICLR 2026Keep the Best, Forget the Rest: Reliable Alignment with Order-Aware Preference Optimization
133. ICLR 2026Landing with the Score: Riemannian Optimization through Denoising
134. ICLR 2026Latent Thinking Optimization: Your Latent Reasoning Language Model Secretly Encodes Reward Signals in its Latent Thoughts
135. ICLR 2026Learning from Noisy Preferences: A Semi-Supervised Learning Approach to Direct Preference Optimization
136. ICLR 2026Learning More with Less: A Dynamic Dual-Level Down-Sampling Framework for Efficient Policy Optimization
137. ICLR 2026Learning of Population Dynamics: Inverse Optimization Meets JKO Scheme
138. ICLR 2026Learning to Segment for Vehicle Routing Problems
139. ICLR 2026Lifelong Learning with Behavior Consolidation for Vehicle Routing
140. ICLR 2026LMask: Learn to Solve Constrained Routing Problems with Lazy Masking
141. ICLR 2026Local Entropy Search over Descent Sequences for Bayesian Optimization
142. ICLR 2026Machine Unlearning under Retain–Forget Entanglement
143. ICLR 2026MaskInversion: Localized Embeddings via Optimization of Explainability Maps
144. ICLR 2026Mitigating Mismatch within Reference-based Preference Optimization
145. ICLR 2026Mitigating Spurious Correlation via Distributionally Robust Learning with Hierarchical Ambiguity Sets
146. ICLR 2026MuonBP: Faster Muon via Block-Periodic Orthogonalization
147. ICLR 2026Native Adaptive Solution Expansion for Diffusion-based Combinatorial Optimization
148. ICLR 2026Nesterov Finds GRAAL: Optimal and Adaptive Gradient Method for Convex Optimization
149. ICLR 2026NeuCLIP: Efficient Large-Scale CLIP Training with Neural Normalizer Optimization
150. ICLR 2026Newton Method Revisited: Global Convergence Rates up to $O(1/k^3)$ for Stepsize Schedules and Linesearch Procedures

常见问题

2023–2026 年四大会议一共有多少篇非凸优化论文？

共 2,330 篇，按年份分别是 2023 年 346 篇、2024 年 560 篇、2025 年 667 篇、2026 年 757 篇。

NeurIPS 2026 有多少篇非凸优化论文？

NeurIPS 2026 有 250 篇非凸优化论文，上一届是 279 篇。

ICLR 2026 有多少篇非凸优化论文？

ICLR 2026 有 178 篇非凸优化论文，上一届是 177 篇。

ICML 2026 有多少篇非凸优化论文？

ICML 2026 有 286 篇非凸优化论文，上一届是 192 篇。

CVPR 2026 有多少篇非凸优化论文？

CVPR 2026 有 43 篇非凸优化论文，上一届是 19 篇。

哪个会议的非凸优化论文最多？

NeurIPS，2023–2026 年共 958 篇。