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 1,339 篇，第 24 / 27 页

  1. ICLR 2024FedImpro: Measuring and Improving Client Update in Federated Learning
  2. ICLR 2024FedLoGe: Joint Local and Generic Federated Learning under Long-tailed Data
  3. ICLR 2024Few-Shot Detection of Machine-Generated Text using Style Representations
  4. ICLR 2024Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!
  5. ICLR 2024Fine-Tuning Language Models for Factuality
  6. ICLR 2024Finetuning Text-to-Image Diffusion Models for Fairness
  7. ICLR 2024FLASK: Fine-grained Language Model Evaluation based on Alignment Skill Sets
  8. ICLR 2024Gaining Wisdom from Setbacks: Aligning Large Language Models via Mistake Analysis
  9. ICLR 2024Generative Judge for Evaluating Alignment
  10. ICLR 2024GPT-4 Is Too Smart To Be Safe: Stealthy Chat with LLMs via Cipher
  11. ICLR 2024Group Preference Optimization: Few-Shot Alignment of Large Language Models
  12. ICLR 2024How to Catch an AI Liar: Lie Detection in Black-Box LLMs by Asking Unrelated Questions
  13. ICLR 2024Identifying the Risks of LM Agents with an LM-Emulated Sandbox
  14. ICLR 2024Implicit Neural Representations and the Algebra of Complex Wavelets
  15. ICLR 2024Improving Generalization of Alignment with Human Preferences through Group Invariant Learning
  16. ICLR 2024Interventional Fairness on Partially Known Causal Graphs: A Constrained Optimization Approach
  17. ICLR 2024Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models
  18. ICLR 2024Language Model Decoding as Direct Metrics Optimization
  19. ICLR 2024Language Model Inversion
  20. ICLR 2024Language-Interfaced Tabular Oversampling via Progressive Imputation and Self-Authentication
  21. ICLR 2024LanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment
  22. ICLR 2024Learning From Simplicial Data Based on Random Walks and 1D Convolutions
  23. ICLR 2024Leftover Lunch: Advantage-based Offline Reinforcement Learning for Language Models
  24. ICLR 2024Leveraging Generative Models for Unsupervised Alignment of Neural Time Series Data
  25. ICLR 2024Like Oil and Water: Group Robustness Methods and Poisoning Defenses May Be at Odds
  26. ICLR 2024LMSYS-Chat-1M: A Large-Scale Real-World LLM Conversation Dataset
  27. ICLR 2024Making RL with Preference-based Feedback Efficient via Randomization
  28. ICLR 2024MAP IT to Visualize Representations
  29. ICLR 2024Masks, Signs, And Learning Rate Rewinding
  30. ICLR 2024MINT: Evaluating LLMs in Multi-turn Interaction with Tools and Language Feedback
  31. ICLR 2024Motif: Intrinsic Motivation from Artificial Intelligence Feedback
  32. ICLR 2024Multilingual Jailbreak Challenges in Large Language Models
  33. ICLR 2024NEFTune: Noisy Embeddings Improve Instruction Finetuning
  34. ICLR 2024Network Memory Footprint Compression Through Jointly Learnable Codebooks and Mappings
  35. ICLR 2024Neuron Activation Coverage: Rethinking Out-of-distribution Detection and Generalization
  36. ICLR 2024Neuron-Enhanced AutoEncoder Matrix Completion and Collaborative Filtering: Theory and Practice
  37. ICLR 2024On Bias-Variance Alignment in Deep Models
  38. ICLR 2024On the Fairness ROAD: Robust Optimization for Adversarial Debiasing
  39. ICLR 2024On the Generalization and Approximation Capacities of Neural Controlled Differential Equations
  40. ICLR 2024On the Over-Memorization During Natural, Robust and Catastrophic Overfitting
  41. ICLR 2024On-Policy Distillation of Language Models: Learning from Self-Generated Mistakes
  42. ICLR 2024Optimal transport based adversarial patch to leverage large scale attack transferability
  43. ICLR 2024Overthinking the Truth: Understanding how Language Models Process False Demonstrations
  44. ICLR 2024PARL: A Unified Framework for Policy Alignment in Reinforcement Learning from Human Feedback
  45. ICLR 2024Plug-and-Play: An Efficient Post-training Pruning Method for Large Language Models
  46. ICLR 2024Privately Aligning Language Models with Reinforcement Learning
  47. ICLR 2024Quantifying and Enhancing Multi-modal Robustness with Modality Preference
  48. ICLR 2024Query-Dependent Prompt Evaluation and Optimization with Offline Inverse RL
  49. ICLR 2024Quick-Tune: Quickly Learning Which Pretrained Model to Finetune and How
  50. ICLR 2024RAIN: Your Language Models Can Align Themselves without Finetuning

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