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 1,421 篇，第 20 / 29 页

  1. NeurIPS 2024AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge Bases
  2. NeurIPS 2024AHA: Human-Assisted Out-of-Distribution Generalization and Detection
  3. NeurIPS 2024ALI-Agent: Assessing LLMs' Alignment with Human Values via Agent-based Evaluation
  4. NeurIPS 2024Aligner-Encoders: Self-Attention Transformers Can Be Self-Transducers
  5. NeurIPS 2024Aligner: Efficient Alignment by Learning to Correct
  6. NeurIPS 2024Aligning Embeddings and Geometric Random Graphs: Informational Results and Computational Approaches for the Procrustes-Wasserstein Problem
  7. NeurIPS 2024Aligning Large Language Models with Representation Editing: A Control Perspective
  8. NeurIPS 2024Aligning Model Properties via Conformal Risk Control
  9. NeurIPS 2024Aligning to Thousands of Preferences via System Message Generalization
  10. NeurIPS 2024Alignment at Pre-training! Towards Native Alignment for Arabic LLMs
  11. NeurIPS 2024Alignment for Honesty
  12. NeurIPS 2024Antigen-Specific Antibody Design via Direct Energy-based Preference Optimization
  13. NeurIPS 2024ART: Automatic Red-teaming for Text-to-Image Models to Protect Benign Users
  14. NeurIPS 2024Ask, Attend, Attack: An Effective Decision-Based Black-Box Targeted Attack for Image-to-Text Models
  15. NeurIPS 2024Axioms for AI Alignment from Human Feedback
  16. NeurIPS 2024BackdoorAlign: Mitigating Fine-tuning based Jailbreak Attack with Backdoor Enhanced Safety Alignment
  17. NeurIPS 2024BAN: Detecting Backdoors Activated by Adversarial Neuron Noise
  18. NeurIPS 2024Bayesian-guided Label Mapping for Visual Reprogramming
  19. NeurIPS 2024Benchmarking PtO and PnO Methods in the Predictive Combinatorial Optimization Regime
  20. NeurIPS 2024Beware of Road Markings: A New Adversarial Patch Attack to Monocular Depth Estimation
  21. NeurIPS 2024Beyond Slow Signs in High-fidelity Model Extraction
  22. NeurIPS 2024Bileve: Securing Text Provenance in Large Language Models Against Spoofing with Bi-level Signature
  23. NeurIPS 2024Biomedical Visual Instruction Tuning with Clinician Preference Alignment
  24. NeurIPS 2024BoNBoN Alignment for Large Language Models and the Sweetness of Best-of-n Sampling
  25. NeurIPS 2024Boosting the Potential of Large Language Models with an Intelligent Information Assistant
  26. NeurIPS 2024Boosting the Transferability of Adversarial Attack on Vision Transformer with Adaptive Token Tuning
  27. NeurIPS 2024Breaking the False Sense of Security in Backdoor Defense through Re-Activation Attack
  28. NeurIPS 2024Cal-DPO: Calibrated Direct Preference Optimization for Language Model Alignment
  29. NeurIPS 2024Calibrated Self-Rewarding Vision Language Models
  30. NeurIPS 2024Can an AI Agent Safely Run a Government? Existence of Probably Approximately Aligned Policies
  31. NeurIPS 2024Can Large Language Model Agents Simulate Human Trust Behavior?
  32. NeurIPS 2024Can Models Learn Skill Composition from Examples?
  33. NeurIPS 2024Can Simple Averaging Defeat Modern Watermarks?
  34. NeurIPS 2024Can Transformers Smell Like Humans?
  35. NeurIPS 2024Catastrophic Goodhart: regularizing RLHF with KL divergence does not mitigate heavy-tailed reward misspecification
  36. NeurIPS 2024CausalDiff: Causality-Inspired Disentanglement via Diffusion Model for Adversarial Defense
  37. NeurIPS 2024Certified Adversarial Robustness via Randomized $\alpha$-Smoothing for Regression Models
  38. NeurIPS 2024Certified Robustness for Deep Equilibrium Models via Serialized Random Smoothing
  39. NeurIPS 2024Chain of Preference Optimization: Improving Chain-of-Thought Reasoning in LLMs
  40. NeurIPS 2024Classifier Clustering and Feature Alignment for Federated Learning under Distributed Concept Drift
  41. NeurIPS 2024CLAVE: An Adaptive Framework for Evaluating Values of LLM Generated Responses
  42. NeurIPS 2024CLIPCEIL: Domain Generalization through CLIP via Channel rEfinement and Image-text aLignment
  43. NeurIPS 2024CLUES: Collaborative Private-domain High-quality Data Selection for LLMs via Training Dynamics
  44. NeurIPS 2024CNCA: Toward Customizable and Natural Generation of Adversarial Camouflage for Vehicle Detectors
  45. NeurIPS 2024Coded Computing for Resilient Distributed Computing: A Learning-Theoretic Framework
  46. NeurIPS 2024ColJailBreak: Collaborative Generation and Editing for Jailbreaking Text-to-Image Deep Generation
  47. NeurIPS 2024CoMat: Aligning Text-to-Image Diffusion Model with Image-to-Text Concept Matching
  48. NeurIPS 2024Confidence Calibration of Classifiers with Many Classes
  49. NeurIPS 2024Conformal Alignment: Knowing When to Trust Foundation Models with Guarantees
  50. NeurIPS 2024Conformal Prediction for Class-wise Coverage via Augmented Label Rank Calibration

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