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 3 / 6 页

101. ICML 2025Gradient Descent Converges Arbitrarily Fast for Logistic Regression via Large and Adaptive Stepsizes
102. ICML 2025GRAIL: Graph Edit Distance and Node Alignment using LLM-Generated Code
103. ICML 2025GRAM: A Generative Foundation Reward Model for Reward Generalization
104. ICML 2025Hardware and Software Platform Inference
105. ICML 2025Hierarchical Graph Tokenization for Molecule-Language Alignment
106. ICML 2025How to set AdamW's weight decay as you scale model and dataset size
107. ICML 2025HPS: Hard Preference Sampling for Human Preference Alignment
108. ICML 2025Human-Aligned Image Models Improve Visual Decoding from the Brain
109. ICML 2025ICLShield: Exploring and Mitigating In-Context Learning Backdoor Attacks
110. ICML 2025Implicit Language Models are RNNs: Balancing Parallelization and Expressivity
111. ICML 2025Improving LLM Safety Alignment with Dual-Objective Optimization
112. ICML 2025Improving Model Alignment Through Collective Intelligence of Open-Source Models
113. ICML 2025Improving Multimodal Learning Balance and Sufficiency through Data Remixing
114. ICML 2025InfAlign: Inference-aware language model alignment
115. ICML 2025Inference-Time Alignment of Diffusion Models with Direct Noise Optimization
116. ICML 2025Invariance Makes LLM Unlearning Resilient Even to Unanticipated Downstream Fine-Tuning
117. ICML 2025Is Best-of-N the Best of Them? Coverage, Scaling, and Optimality in Inference-Time Alignment
118. ICML 2025ITFormer: Bridging Time Series and Natural Language for Multi-Modal QA with Large-Scale Multitask Dataset
119. ICML 2025Jacobian Sparse Autoencoders: Sparsify Computations, Not Just Activations
120. ICML 2025Joint Metric Space Embedding by Unbalanced Optimal Transport with Gromov–Wasserstein Marginal Penalization
121. ICML 2025Just Enough Shifts: Mitigating Over-Refusal in Aligned Language Models with Targeted Representation Fine-Tuning
122. ICML 2025Kernel Quantile Embeddings and Associated Probability Metrics
123. ICML 2025Knowledge Swapping via Learning and Unlearning
124. ICML 2025Knowledge-Guided Wasserstein Distributionally Robust Optimization
125. ICML 2025Language Models as Implicit Tree Search
126. ICML 2025LapSum - One Method to Differentiate Them All: Ranking, Sorting and Top-k Selection
127. ICML 2025Larger or Smaller Reward Margins to Select Preferences for LLM Alignment?
128. ICML 2025Latent Preference Coding: Aligning Large Language Models via Discrete Latent Codes
129. ICML 2025Layer-wise Alignment: Examining Safety Alignment Across Image Encoder Layers in Vision Language Models
130. ICML 2025Learning Along the Arrow of Time: Hyperbolic Geometry for Backward-Compatible Representation Learning
131. ICML 2025Learning Cascade Ranking as One Network
132. ICML 2025Learning from others' mistakes: Finetuning machine translation models with span-level error annotations
133. ICML 2025Learning Safety Constraints for Large Language Models
134. ICML 2025Learning Strategic Language Agents in the Werewolf Game with Iterative Latent Space Policy Optimization
135. ICML 2025Learning the RoPEs: Better 2D and 3D Position Encodings with STRING
136. ICML 2025Learnware Specification via Dual Alignment
137. ICML 2025Leveraging Sparsity for Sample-Efficient Preference Learning: A Theoretical Perspective
138. ICML 2025Linear Bandits with Partially Observable Features
139. ICML 2025LIVS: A Pluralistic Alignment Dataset for Inclusive Public Spaces
140. ICML 2025LLM Alignment as Retriever Optimization: An Information Retrieval Perspective
141. ICML 2025LLM Data Selection and Utilization via Dynamic Bi-level Optimization
142. ICML 2025Logarithmic Regret for Online KL-Regularized Reinforcement Learning
143. ICML 2025LOGO --- Long cOntext aliGnment via efficient preference Optimization
144. ICML 2025Long-Short Alignment for Effective Long-Context Modeling in LLMs
145. ICML 2025LoRA-One: One-Step Full Gradient Could Suffice for Fine-Tuning Large Language Models, Provably and Efficiently
146. ICML 2025M3-JEPA: Multimodal Alignment via Multi-gate MoE based on the Joint-Embedding Predictive Architecture
147. ICML 2025Measuring Diversity: Axioms and Challenges
148. ICML 2025Meta-Reinforcement Learning with Adaptation from Human Feedback via Preference-Order-Preserving Task Embedding
149. ICML 2025MetaOptimize: A Framework for Optimizing Step Sizes and Other Meta-parameters
150. ICML 2025Mind the Gap: a Spectral Analysis of Rank Collapse and Signal Propagation in Attention Layers

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。