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 1,421 篇，第 19 / 29 页

901. NeurIPS 2025Trajectory Balance with Asynchrony: Decoupling Exploration and Learning for Fast, Scalable LLM Post-Training
902. NeurIPS 2025Trajectory Graph Learning: Aligning with Long Trajectories in Reinforcement Learning Without Reward Design
903. NeurIPS 2025TransMLA: Migrating GQA Models to MLA with Full DeepSeek Compatibility and Speedup
904. NeurIPS 2025Triplets Better Than Pairs: Towards Stable and Effective Self-Play Fine-Tuning for LLMs
905. NeurIPS 2025U-REPA: Aligning Diffusion U-Nets to ViTs
906. NeurIPS 2025UMU-Bench: Closing the Modality Gap in Multimodal Unlearning Evaluation
907. NeurIPS 2025Understanding and Rectifying Safety Perception Distortion in VLMs
908. NeurIPS 2025Unifying Attention Heads and Task Vectors via Hidden State Geometry in In-Context Learning
909. NeurIPS 2025Unifying Reconstruction and Density Estimation via Invertible Contraction Mapping in One-Class Classification
910. NeurIPS 2025UniGen: Enhanced Training & Test-Time Strategies for Unified Multimodal Understanding and Generation
911. NeurIPS 2025Unlocker: Disentangle the Deadlock of Learning between Label-noisy and Long-tailed Data
912. NeurIPS 2025Unveiling the Power of Multiple Gossip Steps: A Stability-Based Generalization Analysis in Decentralized Training
913. NeurIPS 2025Unveiling the Uncertainty in Embodied and Operational Carbon of Large AI Models through a Probabilistic Carbon Accounting Model
914. NeurIPS 2025Utility Engineering: Analyzing and Controlling Emergent Value Systems in AIs
915. NeurIPS 2025Value Gradient Guidance for Flow Matching Alignment
916. NeurIPS 2025Vicinal Label Supervision for Reliable Aleatoric and Epistemic Uncertainty Estimation
917. NeurIPS 2025Video-SafetyBench: A Benchmark for Safety Evaluation of Video LVLMs
918. NeurIPS 2025VL-SAE: Interpreting and Enhancing Vision-Language Alignment with a Unified Concept Set
919. NeurIPS 2025VLForgery Face Triad: Detection, Localization and Attribution via Multimodal Large Language Models
920. NeurIPS 2025VLMs can Aggregate Scattered Training Patches
921. NeurIPS 2025VMDT: Decoding the Trustworthiness of Video Foundation Models
922. NeurIPS 2025VPO: Reasoning Preferences Optimization Based on $\mathcal{V}$-Usable Information
923. NeurIPS 2025Walking the Tightrope: Autonomous Disentangling Beneficial and Detrimental Drifts in Non-Stationary Custom-Tuning
924. NeurIPS 2025Weak-to-Strong Generalization under Distribution Shifts
925. NeurIPS 2025WebThinker: Empowering Large Reasoning Models with Deep Research Capability
926. NeurIPS 2025What Makes a Reward Model a Good Teacher? An Optimization Perspective
927. NeurIPS 2025What Matters in Data for DPO?
928. NeurIPS 2025When and how can inexact generative models still sample from the data manifold?
929. NeurIPS 2025When Do Transformers Outperform Feedforward and Recurrent Networks? A Statistical Perspective
930. NeurIPS 2025Who Speaks for the Trigger? Dynamic Expert Routing in Backdoored Mixture-of-Experts Transformers
931. NeurIPS 2025Whose Instructions Count? Resolving Preference Bias in Instruction Fine-Tuning
932. NeurIPS 2025Whose View of Safety? A Deep DIVE Dataset for Pluralistic Alignment of Text-to-Image Models
933. NeurIPS 2025Why and How LLMs Hallucinate: Connecting the Dots with Subsequence Associations
934. NeurIPS 2025Why Do Some Language Models Fake Alignment While Others Don't?
935. NeurIPS 2025Zooming from Context to Cue: Hierarchical Preference Optimization for Multi-Image MLLMs
936. NeurIPS 2024$\beta$-DPO: Direct Preference Optimization with Dynamic $\beta$
937. NeurIPS 20243D Structure Prediction of Atomic Systems with Flow-based Direct Preference Optimization
938. NeurIPS 2024A Critical Evaluation of AI Feedback for Aligning Large Language Models
939. NeurIPS 2024A Novel Benchmark for Decision-Making in Uncertain and Competitive Games
940. NeurIPS 2024A survey and benchmark of high-dimensional Bayesian optimization of discrete sequences
941. NeurIPS 2024A teacher-teacher framework for clinical language representation learning
942. NeurIPS 2024A theoretical case-study of Scalable Oversight in Hierarchical Reinforcement Learning
943. NeurIPS 2024A Theoretical Understanding of Self-Correction through In-context Alignment
944. NeurIPS 2024Accelerating Greedy Coordinate Gradient and General Prompt Optimization via Probe Sampling
945. NeurIPS 2024Achieving Domain-Independent Certified Robustness via Knowledge Continuity
946. NeurIPS 2024AdaNeg: Adaptive Negative Proxy Guided OOD Detection with Vision-Language Models
947. NeurIPS 2024Adaptive Preference Scaling for Reinforcement Learning with Human Feedback
948. NeurIPS 2024Addressing Spatial-Temporal Heterogeneity: General Mixed Time Series Analysis via Latent Continuity Recovery and Alignment
949. NeurIPS 2024Advancing Tool-Augmented Large Language Models: Integrating Insights from Errors in Inference Trees
950. NeurIPS 2024Adversarial Representation Engineering: A General Model Editing Framework for Large Language Models

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。