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 348 篇，第 6 / 7 页

251. NeurIPS 2024Risk-Averse Fine-tuning of Large Language Models
252. NeurIPS 2024Robust Conformal Prediction Using Privileged Information
253. NeurIPS 2024Robust Prompt Optimization for Defending Language Models Against Jailbreaking Attacks
254. NeurIPS 2024Robust Reinforcement Learning from Corrupted Human Feedback
255. NeurIPS 2024RTify: Aligning Deep Neural Networks with Human Behavioral Decisions
256. NeurIPS 2024Rule Based Rewards for Language Model Safety
257. NeurIPS 2024Safe LoRA: The Silver Lining of Reducing Safety Risks when Finetuning Large Language Models
258. NeurIPS 2024SafeSora: Towards Safety Alignment of Text2Video Generation via a Human Preference Dataset
259. NeurIPS 2024Safetywashing: Do AI Safety Benchmarks Actually Measure Safety Progress?
260. NeurIPS 2024SafeWorld: Geo-Diverse Safety Alignment
261. NeurIPS 2024SampDetox: Black-box Backdoor Defense via Perturbation-based Sample Detoxification
262. NeurIPS 2024Scalable Neural Network Verification with Branch-and-bound Inferred Cutting Planes
263. NeurIPS 2024Scaling Laws for Reward Model Overoptimization in Direct Alignment Algorithms
264. NeurIPS 2024Scanning Trojaned Models Using Out-of-Distribution Samples
265. NeurIPS 2024Secret Collusion among AI Agents: Multi-Agent Deception via Steganography
266. NeurIPS 2024Seeing the Image: Prioritizing Visual Correlation by Contrastive Alignment
267. NeurIPS 2024SEEV: Synthesis with Efficient Exact Verification for ReLU Neural Barrier Functions
268. NeurIPS 2024Selective Generation for Controllable Language Models
269. NeurIPS 2024Self-Calibrated Tuning of Vision-Language Models for Out-of-Distribution Detection
270. NeurIPS 2024Self-Calibrating Conformal Prediction
271. NeurIPS 2024Self-Consuming Generative Models with Curated Data Provably Optimize Human Preferences
272. NeurIPS 2024Self-Supervised Alignment with Mutual Information: Learning to Follow Principles without Preference Labels
273. NeurIPS 2024SelfCodeAlign: Self-Alignment for Code Generation
274. NeurIPS 2024Semi-Discrete Optimal Transport: Nearly Minimax Estimation With Stochastic Gradient Descent and Adaptive Entropic Regularization
275. NeurIPS 2024Sequential Harmful Shift Detection Without Labels
276. NeurIPS 2024SeTAR: Out-of-Distribution Detection with Selective Low-Rank Approximation
277. NeurIPS 2024SETBENCH: Assessing the Analytical and Semantic Robustness of Language Models
278. NeurIPS 2024SG-Bench: Evaluating LLM Safety Generalization Across Diverse Tasks and Prompt Types
279. NeurIPS 2024Shadowcast: Stealthy Data Poisoning Attacks Against Vision-Language Models
280. NeurIPS 2024SHDocs: A dataset, benchmark, and method to efficiently generate high-quality, real-world specular highlight data with near-perfect alignment
281. NeurIPS 2024SimPO: Simple Preference Optimization with a Reference-Free Reward
282. NeurIPS 2024Soft Prompt Threats: Attacking Safety Alignment and Unlearning in Open-Source LLMs through the Embedding Space
283. NeurIPS 2024Soft-Label Integration for Robust Toxicity Classification
284. NeurIPS 2024Spectral Editing of Activations for Large Language Model Alignment
285. NeurIPS 2024SpeechAlign: Aligning Speech Generation to Human Preferences
286. NeurIPS 2024SSDM: Scalable Speech Dysfluency Modeling
287. NeurIPS 2024Stability and Generalization of Adversarial Training for Shallow Neural Networks with Smooth Activation
288. NeurIPS 2024Stepwise Alignment for Constrained Language Model Policy Optimization
289. NeurIPS 2024Stress-Testing Capability Elicitation With Password-Locked Models
290. NeurIPS 2024Studying How to Efficiently and Effectively Guide Models with Explanations - A Reproducibility Study
291. NeurIPS 2024T2VSafetyBench: Evaluating the Safety of Text-to-Video Generative Models
292. NeurIPS 2024Tackling Uncertain Correspondences for Multi-Modal Entity Alignment
293. NeurIPS 2024TARP-VP: Towards Evaluation of Transferred Adversarial Robustness and Privacy on Label Mapping Visual Prompting Models
294. NeurIPS 2024Test-Time Adaptation Induces Stronger Accuracy and Agreement-on-the-Line
295. NeurIPS 2024Text-Guided Attention is All You Need for Zero-Shot Robustness in Vision-Language Models
296. NeurIPS 2024TFGDA: Exploring Topology and Feature Alignment in Semi-supervised Graph Domain Adaptation through Robust Clustering
297. NeurIPS 2024The Art of Saying No: Contextual Noncompliance in Language Models
298. NeurIPS 2024The Best of Both Worlds: On the Dilemma of Out-of-distribution Detection
299. NeurIPS 2024The Factorization Curse: Which Tokens You Predict Underlie the Reversal Curse and More
300. NeurIPS 2024The Importance of Online Data: Understanding Preference Fine-tuning via Coverage

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。