对齐与安全论文趋势 2023–2026

主要结论

• 对齐与安全论文从 2023 年的 438 篇增长到 2026 年的 1,882 篇，三年间扩大到约 4.3 倍。
• NeurIPS 是发对齐与安全论文最多的会议，2023–2026 年共 1,421 篇，占四个会议总量的 31%。
• 交叉方向里涨得最快的是社会层面 · safety：同时属于两个主题的论文从 2023 年的 0 篇增至 2026 年的 137 篇。
• 在「对齐与负责任 AI」的 5 个主题里，它的 2026 年论文数排第 1；同组增长最快的社会层面 · safety是 93 倍，对齐与安全是 4.3 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 2026AICrypto: Evaluating Cryptography Capabilities of Large Language Models
• ICLR 2026A Fair Bayesian Inference through Matched Gibbs Posterior
• CVPR 2026Advancing Cancer Prognosis with Hierarchical Fusion of Genomic Proteomic and
• NeurIPS 2025$\Psi$-Sampler: Initial Particle Sampling for SMC-Based Inference-Time Reward Alignment in Score Models
• ICML 2025A Checks-and-Balances Framework for Context-Aware Ethical AI Alignment
• ICLR 2025$\sigma$-zero: Gradient-based Optimization of $\ell_0$-norm Adversarial Examples
• CVPR 2025AesthetiQ: Enhancing Graphic Layout Design via Aesthetic-Aware Preference Alignment of Multi-modal Large Language Models
• NeurIPS 2024A survey and benchmark of high-dimensional Bayesian optimization of discrete sequences
• ICML 2024A Global Geometric Analysis of Maximal Coding Rate Reduction
• ICLR 2024A unique M-pattern for micro-expression spotting in long videos
• CVPR 2024Decompose-and-Compose: A Compositional Approach to Mitigating Spurious Correlation
• NeurIPS 2023(S)GD over Diagonal Linear Networks: Implicit bias, Large Stepsizes and Edge of Stability
• ICML 2023A Robust Optimisation Perspective on Counterexample-Guided Repair of Neural Networks
• ICLR 2023Aligning Model and Macaque Inferior Temporal Cortex Representations Improves Model-to-Human Behavioral Alignment and Adversarial Robustness
• CVPR 2023Adjustment and Alignment for Unbiased Open Set Domain Adaptation
• NeurIPS 2026$\mathcal{P}$Torch: Narrowing the Gap Between Projection and Gradient-Based Learning
• ICML 2026Alignment Pretraining: AI Discourse Causes Self-Fulfilling (Mis)alignment
• ICLR 2026A2D: Any-Order, Any-Step Safety Alignment for Diffusion Language Models
• CVPR 2026AlignPose Generalizable 6D Pose Estimation via Multi-view Feature-metric Alignment

全部对齐与安全论文

共 300 篇，第 2 / 6 页

51. ICML 2025Compelling ReLU Networks to Exhibit Exponentially Many Linear Regions at Initialization and During Training
52. ICML 2025Compressing tree ensembles through Level-wise Optimization and Pruning
53. ICML 2025Computing Optimal Transport Maps and Wasserstein Barycenters Using Conditional Normalizing Flows
54. ICML 2025Conformal Tail Risk Control for Large Language Model Alignment
55. ICML 2025ConfPO: Exploiting Policy Model Confidence for Critical Token Selection in Preference Optimization
56. ICML 2025Constrain Alignment with Sparse Autoencoders
57. ICML 2025Contextual Optimization Under Model Misspecification: A Tractable and Generalizable Approach
58. ICML 2025Counting in Small Transformers: The Delicate Interplay between Attention and Feed-Forward Layers
59. ICML 2025Cross-City Latent Space Alignment for Consistency Region Embedding
60. ICML 2025CSV-Occ: Fusing Multi-frame Alignment for Occupancy Prediction with Temporal Cross State Space Model and Central Voting Mechanism
61. ICML 2025DAMA: Data- and Model-aware Alignment of Multi-modal LLMs
62. ICML 2025DANCE: Dual Unbiased Expansion with Group-acquired Alignment for Out-of-distribution Graph Fairness Learning
63. ICML 2025Deep Linear Network Training Dynamics from Random Initialization: Data, Width, Depth, and Hyperparameter Transfer
64. ICML 2025Defending LVLMs Against Vision Attacks Through Partial-Perception Supervision
65. ICML 2025Design Considerations in Offline Preference-based RL
66. ICML 2025Differentiable Quadratic Optimization For the Maximum Independent Set Problem
67. ICML 2025Direct Density Ratio Optimization: A Statistically Consistent Approach to Aligning Large Language Models
68. ICML 2025Direct Discriminative Optimization: Your Likelihood-Based Visual Generative Model is Secretly a GAN Discriminator
69. ICML 2025Discrete and Continuous Difference of Submodular Minimization
70. ICML 2025Discriminative Finetuning of Generative Large Language Models without Reward Models and Human Preference Data
71. ICML 2025Distributed Retraction-Free and Communication-Efficient Optimization on the Stiefel Manifold
72. ICML 2025Diverging Preferences: When do Annotators Disagree and do Models Know?
73. ICML 2025Don't Restart, Just Reuse: Reoptimizing MILPs with Dynamic Parameters
74. ICML 2025DPO Meets PPO: Reinforced Token Optimization for RLHF
75. ICML 2025DreamDPO: Aligning Text-to-3D Generation with Human Preferences via Direct Preference Optimization
76. ICML 2025Edge-Colored Clustering in Hypergraphs: Beyond Minimizing Unsatisfied Edges
77. ICML 2025Eigen Analysis of Conjugate Kernel and Neural Tangent Kernel
78. ICML 2025Eliciting Language Model Behaviors with Investigator Agents
79. ICML 2025ELITE: Enhanced Language-Image Toxicity Evaluation for Safety
80. ICML 2025ELMO : Efficiency via Low-precision and Peak Memory Optimization in Large Output Spaces
81. ICML 2025Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs
82. ICML 2025Energy-Based Preference Model Offers Better Offline Alignment than the Bradley-Terry Preference Model
83. ICML 2025Enhancing Ligand Validity and Affinity in Structure-Based Drug Design with Multi-Reward Optimization
84. ICML 2025Enhancing Parallelism in Decentralized Stochastic Convex Optimization
85. ICML 2025Enhancing Spectral GNNs: From Topology and Perturbation Perspectives
86. ICML 2025EquivaMap: Leveraging LLMs for Automatic Equivalence Checking of Optimization Formulations
87. ICML 2025EraseAnything: Enabling Concept Erasure in Rectified Flow Transformers
88. ICML 2025Evolving Prompts In-Context: An Open-ended, Self-replicating Perspective
89. ICML 2025Exact risk curves of signSGD in High-Dimensions: quantifying preconditioning and noise-compression effects
90. ICML 2025Explicit Preference Optimization: No Need for an Implicit Reward Model
91. ICML 2025Faster Global Minimum Cut with Predictions
92. ICML 2025Features are fate: a theory of transfer learning in high-dimensional regression
93. ICML 2025FlipAttack: Jailbreak LLMs via Flipping
94. ICML 2025Flow Matching for Few-Trial Neural Adaptation with Stable Latent Dynamics
95. ICML 2025Functional Alignment Can Mislead: Examining Model Stitching
96. ICML 2025Fusing Reward and Dueling Feedback in Stochastic Bandits
97. ICML 2025Gandalf the Red: Adaptive Security for LLMs
98. ICML 2025Generalizing from SIMPLE to HARD Visual Reasoning: Can We Mitigate Modality Imbalance in VLMs?
99. ICML 2025Global Convergence and Rich Feature Learning in $L$-Layer Infinite-Width Neural Networks under $\mu$ Parametrization
100. ICML 2025Global Optimization with a Power-Transformed Objective and Gaussian Smoothing

常见问题

2023–2026 年四大会议一共有多少篇对齐与安全论文？

共 4,558 篇，按年份分别是 2023 年 438 篇、2024 年 875 篇、2025 年 1,363 篇、2026 年 1,882 篇。

NeurIPS 2026 有多少篇对齐与安全论文？

NeurIPS 2026 有 480 篇对齐与安全论文，上一届是 455 篇。

ICLR 2026 有多少篇对齐与安全论文？

ICLR 2026 有 623 篇对齐与安全论文，上一届是 473 篇。

ICML 2026 有多少篇对齐与安全论文？

ICML 2026 有 551 篇对齐与安全论文，上一届是 300 篇。

CVPR 2026 有多少篇对齐与安全论文？

CVPR 2026 有 228 篇对齐与安全论文，上一届是 135 篇。

哪个会议的对齐与安全论文最多？

NeurIPS，2023–2026 年共 1,421 篇。