对齐与安全论文趋势 2023–2026

主要结论

• 对齐与安全论文从 2023 年的 438 篇增长到 2026 年的 1,882 篇，三年间扩大到约 4.3 倍。
• NeurIPS 是发对齐与安全论文最多的会议，2023–2026 年共 1,421 篇，占四个会议总量的 31%。
• 交叉方向里涨得最快的是社会层面 · safety：同时属于两个主题的论文从 2023 年的 0 篇增至 2026 年的 137 篇。
• 在「对齐与负责任 AI」的 5 个主题里，它的 2026 年论文数排第 1；同组增长最快的社会层面 · safety是 93 倍，对齐与安全是 4.3 倍。

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

• NeurIPS 2026$\gamma$-weakly $\theta$-up-concavity: A Unified Framework for Non-Convex Optimization Beyond DR-Submodular and OSS Functions
• ICML 2026AICrypto: Evaluating Cryptography Capabilities of Large Language Models
• ICLR 2026A Fair Bayesian Inference through Matched Gibbs Posterior
• CVPR 2026Advancing Cancer Prognosis with Hierarchical Fusion of Genomic Proteomic and
• NeurIPS 2025$\Psi$-Sampler: Initial Particle Sampling for SMC-Based Inference-Time Reward Alignment in Score Models
• ICML 2025A Checks-and-Balances Framework for Context-Aware Ethical AI Alignment
• ICLR 2025$\sigma$-zero: Gradient-based Optimization of $\ell_0$-norm Adversarial Examples
• CVPR 2025AesthetiQ: Enhancing Graphic Layout Design via Aesthetic-Aware Preference Alignment of Multi-modal Large Language Models
• NeurIPS 2024A survey and benchmark of high-dimensional Bayesian optimization of discrete sequences
• ICML 2024A Global Geometric Analysis of Maximal Coding Rate Reduction
• ICLR 2024A unique M-pattern for micro-expression spotting in long videos
• CVPR 2024Decompose-and-Compose: A Compositional Approach to Mitigating Spurious Correlation
• NeurIPS 2023(S)GD over Diagonal Linear Networks: Implicit bias, Large Stepsizes and Edge of Stability
• ICML 2023A Robust Optimisation Perspective on Counterexample-Guided Repair of Neural Networks
• ICLR 2023Aligning Model and Macaque Inferior Temporal Cortex Representations Improves Model-to-Human Behavioral Alignment and Adversarial Robustness
• CVPR 2023Adjustment and Alignment for Unbiased Open Set Domain Adaptation
• NeurIPS 2026$\mathcal{P}$Torch: Narrowing the Gap Between Projection and Gradient-Based Learning
• ICML 2026Alignment Pretraining: AI Discourse Causes Self-Fulfilling (Mis)alignment
• ICLR 2026A2D: Any-Order, Any-Step Safety Alignment for Diffusion Language Models
• CVPR 2026AlignPose Generalizable 6D Pose Estimation via Multi-view Feature-metric Alignment

全部对齐与安全论文

共 4,558 篇，第 14 / 92 页

651. ICLR 2026Private Rate-Constrained Optimization with Applications to Fair Learning
652. ICLR 2026PromptHub: Enhancing Multi-Prompt Visual In-Context Learning with Locality-Aware Fusion, Concentration and Alignment
653. ICLR 2026Propaganda AI: An Analysis of Semantic Divergence in Large Language Models
654. ICLR 2026PropensityBench: Evaluating Latent Safety Risks in Large Language Models via an Agentic Approach
655. ICLR 2026Property-Driven Protein Inverse Folding with Multi-Objective Preference Alignment
656. ICLR 2026ProSafePrune: Projected Safety Pruning for Mitigating Over-Refusal in LLMs
657. ICLR 2026Protection against Source Inference Attacks in Federated Learning
658. ICLR 2026Pruning Long Chain-of-Thought of Large Reasoning Models via Small-Scale Preference Optimization
659. ICLR 2026REAL: Reading Out Transformer Activations for Precise Localization in Language Model Steering
660. ICLR 2026Reasoned Safety Alignment: Ensuring Jailbreak Defense via Answer-Then-Check
661. ICLR 2026Reasoning Boosts Opinion Alignment in LLMs
662. ICLR 2026RedCodeAgent: Automatic Red-teaming Agent against Diverse Code Agents
663. ICLR 2026RedTeamCUA: Realistic Adversarial Testing of Computer-Use Agents in Hybrid Web-OS Environments
664. ICLR 2026Reducing information dependency does not cause training data privacy. Adversarially non-robust features do.
665. ICLR 2026References Improve LLM Alignment in Non-Verifiable Domains
666. ICLR 2026Regulating Internal Evidence Flows for Robust Learning Under Spurious Correlations
667. ICLR 2026Reinforcement Learning Fine-Tuning Enhances Activation Intensity and Diversity in the Internal Circuitry of LLMs
668. ICLR 2026Reinforcement Unlearning via Group Relative Policy Optimization
669. ICLR 2026Relationship Alignment for View-aware Multi-view Clustering
670. ICLR 2026RelayFormer: A Unified Local-Global Attention Framework for Scalable Image and Video Manipulation Localization
671. ICLR 2026Reliable Poisoned Sample Detection against Backdoor Attacks Enhanced by Sharpness Aware Minimization
672. ICLR 2026Reliable Weak-to-Strong Monitoring of LLM Agents
673. ICLR 2026RepIt: Steering Language Models with Concept-Specific Refusal Vectors
674. ICLR 2026Representation Alignment for Diffusion Transformers without External Components
675. ICLR 2026Representational Alignment Across Model Layers and Brain Regions with Hierarchical Optimal Transport
676. ICLR 2026RESFL: An Uncertainty-Aware Framework for Responsible Federated Learning by Balancing Privacy, Fairness and Utility
677. ICLR 2026Rethinking Benign Relearning: Syntax as the Hidden Driver of Unlearning Failures
678. ICLR 2026Rethinking Bottlenecks in Safety Fine-Tuning of Vision Language Models
679. ICLR 2026Rethinking LoRA for Privacy-Preserving Federated Learning in Large Models
680. ICLR 2026Rethinking Pareto Frontier: On the Optimal Trade-offs in Fair Classification
681. ICLR 2026Rethinking Radiology Report Generation: From Narrative Flow to Topic-Guided Findings
682. ICLR 2026Rethinking the Gold Standard: Why Discrete Curvature Fails to Fully Capture Over-squashing in GNNs?
683. ICLR 2026Reward Model Routing in Alignment
684. ICLR 2026Reward Models Inherit Value Biases from Pretraining
685. ICLR 2026RewardEval: Advancing Reward Model Evaluation
686. ICLR 2026Riemannian Optimization on Relaxed Indicator Matrix Manifold
687. ICLR 2026Risk Phase Transitions in Spiked Regression: Alignment Driven Benign and Catastrophic Overfitting
688. ICLR 2026Risk-Sensitive Agent Compositions
689. ICLR 2026RLBFF: Binary Flexible Feedback to bridge between Human Feedback & Verifiable Rewards
690. ICLR 2026Robust Adversarial Attacks Against Unknown Disturbance via Inverse Gradient Sample
691. ICLR 2026Robust LLM Unlearning via Post Judgment and Multi-round Thinking
692. ICLR 2026Robust Multi-Objective Controlled Decoding of Large Language Models
693. ICLR 2026Robust Preference Alignment via Directional Neighborhood Consensus
694. ICLR 2026Robust Preference Optimization: Aligning Language Models with Noisy Preference Feedback
695. ICLR 2026Robust Reward Modeling via Causal Rubrics
696. ICLR 2026Routing Manifold Alignment Improves Generalization of Mixture-of-Experts LLMs
697. ICLR 2026Routing, Cascades, and User Choice for LLMs
698. ICLR 2026Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domains
699. ICLR 2026SafeDialBench: A Fine-Grained Safety Evaluation Benchmark for Large Language Models in Multi-Turn Dialogues with Diverse Jailbreak Attacks
700. ICLR 2026SafeDPO: A Simple Approach to Direct Preference Optimization with Enhanced Safety

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。