对齐与安全论文趋势 2023–2026

RLHF、偏好优化、红队与安全对齐

统计范围：NeurIPS、ICLR、ICML、CVPR 四个会议，2023–2026 年，共 4,558 篇被归入该主题的论文。

主要结论

每年论文数

各会议论文数

按年份和会议统计

对齐与安全论文数，按年份和会议
年份NeurIPSICLRICMLCVPR合计
202313887101112438
2024348156246125875
20254554733001351,363
20264806235512281,882

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

全部对齐与安全论文

年份全部2023202420252026
会议全部NeurIPSICLRICMLCVPR

共 875 篇，第 7 / 18 页

  1. ICML 2024Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications
  2. ICML 2024AutoOS: Make Your OS More Powerful by Exploiting Large Language Models
  3. ICML 2024Balanced Resonate-and-Fire Neurons
  4. ICML 2024Balancing Similarity and Complementarity for Federated Learning
  5. ICML 2024Bootstrapping Fisher Market Equilibrium and First-Price Pacing Equilibrium
  6. ICML 2024BRAIn: Bayesian Reward-conditioned Amortized Inference for natural language generation from feedback
  7. ICML 2024Bridging Mini-Batch and Asymptotic Analysis in Contrastive Learning: From InfoNCE to Kernel-Based Losses
  8. ICML 2024Byzantine-Robust Federated Learning: Impact of Client Subsampling and Local Updates
  9. ICML 2024Can AI Assistants Know What They Don't Know?
  10. ICML 2024Can Machines Learn the True Probabilities?
  11. ICML 2024CARTE: Pretraining and Transfer for Tabular Learning
  12. ICML 2024Catapults in SGD: spikes in the training loss and their impact on generalization through feature learning
  13. ICML 2024Causal Bandits: The Pareto Optimal Frontier of Adaptivity, a Reduction to Linear Bandits, and Limitations around Unknown Marginals
  14. ICML 2024Characterizing Large Language Model Geometry Helps Solve Toxicity Detection and Generation
  15. ICML 2024Classification under Nuisance Parameters and Generalized Label Shift in Likelihood-Free Inference
  16. ICML 2024CogBench: a large language model walks into a psychology lab
  17. ICML 2024COLD-Attack: Jailbreaking LLMs with Stealthiness and Controllability
  18. ICML 2024Compressing Large Language Models by Joint Sparsification and Quantization
  19. ICML 2024Conformalized Survival Distributions: A Generic Post-Process to Increase Calibration
  20. ICML 2024Contrastive Predict-and-Search for Mixed Integer Linear Programs
  21. ICML 2024Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation
  22. ICML 2024Convergence and Complexity Guarantee for Inexact First-order Riemannian Optimization Algorithms
  23. ICML 2024Convergence and Trade-Offs in Riemannian Gradient Descent and Riemannian Proximal Point
  24. ICML 2024Convex and Bilevel Optimization for Neural-Symbolic Inference and Learning
  25. ICML 2024Criterion Collapse and Loss Distribution Control
  26. ICML 2024DecisionNCE: Embodied Multimodal Representations via Implicit Preference Learning
  27. ICML 2024Decoding-time Realignment of Language Models
  28. ICML 2024Decomposable Submodular Maximization in Federated Setting
  29. ICML 2024Deconstructing the Goldilocks Zone of Neural Network Initialization
  30. ICML 2024Decoupling Learning and Decision-Making: Breaking the $\mathcal{O}(\sqrt{T})$ Barrier in Online Resource Allocation with First-Order Methods
  31. ICML 2024Deep Networks Always Grok and Here is Why
  32. ICML 2024Defense against Backdoor Attack on Pre-trained Language Models via Head Pruning and Attention Normalization
  33. ICML 2024Delaunay Graph: Addressing Over-Squashing and Over-Smoothing Using Delaunay Triangulation
  34. ICML 2024Dense Reward for Free in Reinforcement Learning from Human Feedback
  35. ICML 2024Detecting and Identifying Selection Structure in Sequential Data
  36. ICML 2024Differentiability and Optimization of Multiparameter Persistent Homology
  37. ICML 2024Differentiable Annealed Importance Sampling Minimizes The Jensen-Shannon Divergence Between Initial and Target Distribution
  38. ICML 2024Distributed High-Dimensional Quantile Regression: Estimation Efficiency and Support Recovery
  39. ICML 2024Distribution Alignment Optimization through Neural Collapse for Long-tailed Classification
  40. ICML 2024Diving into Underwater: Segment Anything Model Guided Underwater Salient Instance Segmentation and A Large-scale Dataset
  41. ICML 2024Do Models Explain Themselves? Counterfactual Simulatability of Natural Language Explanations
  42. ICML 2024EE-LLM: Large-Scale Training and Inference of Early-Exit Large Language Models with 3D Parallelism
  43. ICML 2024Efficient Algorithms for Empirical Group Distributionally Robust Optimization and Beyond
  44. ICML 2024Emergent Equivariance in Deep Ensembles
  45. ICML 2024Energy-based Backdoor Defense without Task-Specific Samples and Model Retraining
  46. ICML 2024Enhancing Implicit Shape Generators Using Topological Regularizations
  47. ICML 2024Enhancing Storage and Computational Efficiency in Federated Multimodal Learning for Large-Scale Models
  48. ICML 2024Equivariant Deep Weight Space Alignment
  49. ICML 2024Equivariant Frames and the Impossibility of Continuous Canonicalization
  50. ICML 2024Estimating Barycenters of Distributions with Neural Optimal Transport

在 Claude、Cursor 里直接问这个趋势

接上 Paperlist MCP 后，写综述、挑方向或准备投稿时，不用离开对话框就能拿到这页的数字：每年各会议的论文数、增长倍数、涨得最快的子方向和代表论文，AI 会按你的问题整理。

可以直接复制这句话问 AI

用 paperlist 查对齐与安全在 NeurIPS 2023–2026 的论文趋势，哪一年涨得最快？

AI 会调用的工具：get_topic_stats {"topic": "alignment-and-safety"}

两分钟接入 Paperlist MCP（Claude Code、Cursor、Codex、ChatGPT）

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。

「对齐与负责任 AI」下的其他主题