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 76 / 92 页

  1. ICML 2024Token-Specific Watermarking with Enhanced Detectability and Semantic Coherence for Large Language Models
  2. ICML 2024Topological Neural Networks go Persistent, Equivariant, and Continuous
  3. ICML 2024Towards AutoAI: Optimizing a Machine Learning System with Black-box and Differentiable Components
  4. ICML 2024Towards Efficient Exact Optimization of Language Model Alignment
  5. ICML 2024Towards Efficient Spiking Transformer: a Token Sparsification Framework for Training and Inference Acceleration
  6. ICML 2024Towards Modular LLMs by Building and Reusing a Library of LoRAs
  7. ICML 2024Towards Scalable and Versatile Weight Space Learning
  8. ICML 2024Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Duality
  9. ICML 2024Transformers, parallel computation, and logarithmic depth
  10. ICML 2024Transforming and Combining Rewards for Aligning Large Language Models
  11. ICML 2024Tuning-free Estimation and Inference of Cumulative Distribution Function under Local Differential Privacy
  12. ICML 2024Understanding MLP-Mixer as a wide and sparse MLP
  13. ICML 2024Understanding the Learning Dynamics of Alignment with Human Feedback
  14. ICML 2024Universality of Linear Recurrences Followed by Non-linear Projections: Finite-Width Guarantees and Benefits of Complex Eigenvalues
  15. ICML 2024Vector Quantization Pretraining for EEG Time Series with Random Projection and Phase Alignment
  16. ICML 2024Verifying message-passing neural networks via topology-based bounds tightening
  17. ICML 2024WARM: On the Benefits of Weight Averaged Reward Models
  18. ICML 2024Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision
  19. ICML 2024Weakly Convex Regularisers for Inverse Problems: Convergence of Critical Points and Primal-Dual Optimisation
  20. ICML 2024What Improves the Generalization of Graph Transformers? A Theoretical Dive into the Self-attention and Positional Encoding
  21. ICML 2024When Linear Attention Meets Autoregressive Decoding: Towards More Effective and Efficient Linearized Large Language Models
  22. ICML 2024When Will Gradient Regularization Be Harmful?
  23. NeurIPS 2024$\beta$-DPO: Direct Preference Optimization with Dynamic $\beta$
  24. NeurIPS 20243D Structure Prediction of Atomic Systems with Flow-based Direct Preference Optimization
  25. NeurIPS 2024A Critical Evaluation of AI Feedback for Aligning Large Language Models
  26. NeurIPS 2024A Novel Benchmark for Decision-Making in Uncertain and Competitive Games
  27. NeurIPS 2024A survey and benchmark of high-dimensional Bayesian optimization of discrete sequences
  28. NeurIPS 2024A teacher-teacher framework for clinical language representation learning
  29. NeurIPS 2024A theoretical case-study of Scalable Oversight in Hierarchical Reinforcement Learning
  30. NeurIPS 2024A Theoretical Understanding of Self-Correction through In-context Alignment
  31. NeurIPS 2024Accelerating Greedy Coordinate Gradient and General Prompt Optimization via Probe Sampling
  32. NeurIPS 2024Achieving Domain-Independent Certified Robustness via Knowledge Continuity
  33. NeurIPS 2024AdaNeg: Adaptive Negative Proxy Guided OOD Detection with Vision-Language Models
  34. NeurIPS 2024Adaptive Preference Scaling for Reinforcement Learning with Human Feedback
  35. NeurIPS 2024Addressing Spatial-Temporal Heterogeneity: General Mixed Time Series Analysis via Latent Continuity Recovery and Alignment
  36. NeurIPS 2024Advancing Tool-Augmented Large Language Models: Integrating Insights from Errors in Inference Trees
  37. NeurIPS 2024Adversarial Representation Engineering: A General Model Editing Framework for Large Language Models
  38. NeurIPS 2024AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge Bases
  39. NeurIPS 2024AHA: Human-Assisted Out-of-Distribution Generalization and Detection
  40. NeurIPS 2024ALI-Agent: Assessing LLMs' Alignment with Human Values via Agent-based Evaluation
  41. NeurIPS 2024Aligner-Encoders: Self-Attention Transformers Can Be Self-Transducers
  42. NeurIPS 2024Aligner: Efficient Alignment by Learning to Correct
  43. NeurIPS 2024Aligning Embeddings and Geometric Random Graphs: Informational Results and Computational Approaches for the Procrustes-Wasserstein Problem
  44. NeurIPS 2024Aligning Large Language Models with Representation Editing: A Control Perspective
  45. NeurIPS 2024Aligning Model Properties via Conformal Risk Control
  46. NeurIPS 2024Aligning to Thousands of Preferences via System Message Generalization
  47. NeurIPS 2024Alignment at Pre-training! Towards Native Alignment for Arabic LLMs
  48. NeurIPS 2024Alignment for Honesty
  49. NeurIPS 2024Antigen-Specific Antibody Design via Direct Energy-based Preference Optimization
  50. NeurIPS 2024ART: Automatic Red-teaming for Text-to-Image Models to Protect Benign Users

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