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 4,558 篇，第 87 / 92 页

  1. ICLR 2023Symmetries, Flat Minima, and the Conserved Quantities of Gradient Flow
  2. ICLR 2023SYNC: SAFETY-AWARE NEURAL CONTROL FOR STABILIZING STOCHASTIC DELAY-DIFFERENTIAL EQUATIONS
  3. ICLR 2023TANGOS: Regularizing Tabular Neural Networks through Gradient Orthogonalization and Specialization
  4. ICLR 2023Targeted Hyperparameter Optimization with Lexicographic Preferences Over Multiple Objectives
  5. ICLR 2023Task Ambiguity in Humans and Language Models
  6. ICLR 2023TempCLR: Temporal Alignment Representation with Contrastive Learning
  7. ICLR 2023Thalamus: a brain-inspired algorithm for biologically-plausible continual learning and disentangled representations
  8. ICLR 2023Towards One-shot Neural Combinatorial Solvers: Theoretical and Empirical Notes on the Cardinality-Constrained Case
  9. ICLR 2023Towards Understanding and Mitigating Dimensional Collapse in Heterogeneous Federated Learning
  10. ICLR 2023Towards Understanding Ensemble, Knowledge Distillation and Self-Distillation in Deep Learning
  11. ICLR 2023Training language models to summarize narratives improves brain alignment
  12. ICLR 2023Understanding Influence Functions and Datamodels via Harmonic Analysis
  13. ICLR 2023Unified Detoxifying and Debiasing in Language Generation via Inference-time Adaptive Optimization
  14. ICLR 2023Unsupervised Manifold Alignment with Joint Multidimensional Scaling
  15. ICLR 2023Voxurf: Voxel-based Efficient and Accurate Neural Surface Reconstruction
  16. ICLR 2023What Is Missing in IRM Training and Evaluation? Challenges and Solutions
  17. ICLR 2023When to Make and Break Commitments?
  18. ICLR 2023Where to Begin? On the Impact of Pre-Training and Initialization in Federated Learning
  19. ICLR 2023Why (and When) does Local SGD Generalize Better than SGD?
  20. ICML 2023A General Theory for Federated Optimization with Asynchronous and Heterogeneous Clients Updates
  21. ICML 2023A Modern Look at the Relationship between Sharpness and Generalization
  22. ICML 2023A Robust Optimisation Perspective on Counterexample-Guided Repair of Neural Networks
  23. ICML 2023A Study on Transformer Configuration and Training Objective
  24. ICML 2023A Theoretical Analysis of the Learning Dynamics under Class Imbalance
  25. ICML 2023abess: A Fast Best-Subset Selection Library in Python and R
  26. ICML 2023Accelerated Infeasibility Detection of Constrained Optimization and Fixed-Point Iterations
  27. ICML 2023Adversarial Classification: Necessary Conditions and Geometric Flows
  28. ICML 2023Aligning Language Models with Preferences through $f$-divergence Minimization
  29. ICML 2023Analysis of Error Feedback in Federated Non-Convex Optimization with Biased Compression: Fast Convergence and Partial Participation
  30. ICML 2023Are Gaussian Data All You Need? The Extents and Limits of Universality in High-Dimensional Generalized Linear Estimation
  31. ICML 2023Are labels informative in semi-supervised learning? Estimating and leveraging the missing-data mechanism.
  32. ICML 2023Bandit Online Linear Optimization with Hints and Queries
  33. ICML 2023Bayesian Neural Networks Avoid Encoding Complex and Perturbation-Sensitive Concepts
  34. ICML 2023Biases in Evaluation of Molecular Optimization Methods and Bias Reduction Strategies
  35. ICML 2023Constrained Optimization via Exact Augmented Lagrangian and Randomized Iterative Sketching
  36. ICML 2023Constrained Phi-Equilibria
  37. ICML 2023Contextual Conservative Interleaving Bandits
  38. ICML 2023Contextual Reliability: When Different Features Matter in Different Contexts
  39. ICML 2023Convex Geometry of ReLU-layers, Injectivity on the Ball and Local Reconstruction
  40. ICML 2023Critical Points and Convergence Analysis of Generative Deep Linear Networks Trained with Bures-Wasserstein Loss
  41. ICML 2023Cut your Losses with Squentropy
  42. ICML 2023Do the Rewards Justify the Means? Measuring Trade-Offs Between Rewards and Ethical Behavior in the Machiavelli Benchmark
  43. ICML 2023DualHSIC: HSIC-Bottleneck and Alignment for Continual Learning
  44. ICML 2023DUET: 2D Structured and Approximately Equivariant Representations
  45. ICML 2023EF21-P and Friends: Improved Theoretical Communication Complexity for Distributed Optimization with Bidirectional Compression
  46. ICML 2023Efficient Sequence Transduction by Jointly Predicting Tokens and Durations
  47. ICML 2023Eliminating Adversarial Noise via Information Discard and Robust Representation Restoration
  48. ICML 2023Expected Gradients of Maxout Networks and Consequences to Parameter Initialization
  49. ICML 2023Explainable Data-Driven Optimization: From Context to Decision and Back Again
  50. ICML 2023FAIRER: Fairness as Decision Rationale Alignment

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