社会层面 · safety论文趋势 2023–2026

主要结论

• 社会层面 · safety论文从 2023 年的 4 篇增长到 2026 年的 370 篇，三年间扩大到约 93 倍。
• ICML 是发社会层面 · safety论文最多的会议，2023–2026 年共 364 篇，占四个会议总量的 53%。
• 交叉方向里涨得最快的是对齐与安全：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 3；同组增长最快的社会层面 · privacy是 6.6 倍，社会层面 · safety是 93 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 2026BarrierSteer: LLM Safety via Learning Barrier Steering
• ICML 2026Anti-Backdoor Coreset Selection via Cumulative Entropy
• ICLR 2026Bi-Criteria Metric Distortion
• CVPR 2026Omni-MMSI Toward Identity-attributed Social Interaction Understanding
• NeurIPS 2025Safety Pretraining: Toward the Next Generation of Safe AI
• ICML 2025$S^2$FGL: Spatial Spectral Federated Graph Learning
• ICLR 2025The Hidden Cost of Waiting for Accurate Predictions
• CVPR 2025Do We Really Need Curated Malicious Data for Safety Alignment in Multi-modal Large Language Models?
• NeurIPS 2024Quantifying the Bitter Lesson: How Safety Benchmarks Measure Capabilities Instead of Safety
• ICML 2024Optimally Improving Cooperative Learning in a Social Setting
• ICLR 2024The Devil is in the Neurons: Interpreting and Mitigating Social Biases in Language Models
• CVPR 2024Digital Life Project: Autonomous 3D Characters with Social Intelligence
• NeurIPS 2023Strategyproof Voting under Correlated Beliefs
• ICML 2023Adaptive Coordination in Social Embodied Rearrangement
• NeurIPS 2026Beyond Feature Disruption: Boundary-Diverting Unlearnable Examples against Linear Probing
• ICML 2026AutoControl Arena: Synthesizing Executable Test Environments for Frontier AI Risk Evaluation
• ICLR 2026PRISON: Unmasking the Criminal Potential of Large Language Models
• CVPR 2026Spot The Ball A Benchmark for Visual Social Inference
• NeurIPS 2025SAFEx: Analyzing Vulnerabilities of MoE-Based LLMs via Stable Safety-critical Expert Identification
• ICML 2025Adversaries Can Misuse Combinations of Safe Models

全部社会层面 · safety论文

共 121 篇，第 3 / 3 页

101. NeurIPS 2026TailGuard: Subgroup Tail Coverage Theory for Safe LLM Alignment
102. NeurIPS 2026Tatemae: Detecting Alignment Faking via Tool Selection in LLMs
103. NeurIPS 2026Tcell: Mitigating Harmful Fine-tuning for Large Language Models via Gradient Alignment
104. NeurIPS 2026Temporal Gradient Inversion for Private Trajectory Reconstruction in Embodied Reinforcement Learning
105. NeurIPS 2026TensorCommitments: A Lightweight Verifiable Inference for Large Language Models
106. NeurIPS 2026The Algorithm Is Not the Behavior: Learned Priors Override Look-Ahead in a Chess-Playing Neural Network
107. NeurIPS 2026The Alignment Tax Concentrates in Output Projections
108. NeurIPS 2026The Geometric Inductive Bias of Grokking: Bypassing Phase Transitions via Architectural Topology
109. NeurIPS 2026The Geometric Wall: Manifold Structure Predicts Layerwise Sparse Autoencoder Scaling Laws
110. NeurIPS 2026The Geometry of Alignment Collapse: When Fine-Tuning Breaks Safety
111. NeurIPS 2026Tight PAC-Bayes Generalisation Guarantees for Large Language Model Safety Monitoring
112. NeurIPS 2026Towards Fair Graph Generation Without Sensitive Attribute
113. NeurIPS 2026Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations
114. NeurIPS 2026Training ML Models with Predictable Failures
115. NeurIPS 2026TrustMod-SM: A Multi-Axis Benchmark for Evaluating Trustworthiness of LLMs in Social Media Content Moderation
116. NeurIPS 2026Unsupervised Unlearnable Segmentation via Semantic Structure Disruption
117. NeurIPS 2026Untrusted Content Masking for Web Agents with Security Guarantees
118. NeurIPS 2026UpSafe℃: Upcycling for Controllable Safety in Large Language Models
119. NeurIPS 2026Values Are Not Single Labels: Distributional Value Profiling Across Groups and Contexts
120. NeurIPS 2026When Poison Meets Structure: Topology-based Defense against Poisoning Attack on Graph-based Retrieval-Augmented Generation
121. NeurIPS 2026Where Should Society Draw the Line? A Social Choice Approach to Collective Consent

常见问题

2023–2026 年四大会议一共有多少篇社会层面 · safety论文？

共 688 篇，按年份分别是 2023 年 4 篇、2024 年 41 篇、2025 年 273 篇、2026 年 370 篇。

NeurIPS 2026 有多少篇社会层面 · safety论文？

NeurIPS 2026 有 121 篇社会层面 · safety论文，上一届是 131 篇。

ICLR 2026 有多少篇社会层面 · safety论文？

ICLR 2026 有 20 篇社会层面 · safety论文，上一届是 11 篇。

ICML 2026 有多少篇社会层面 · safety论文？

ICML 2026 有 226 篇社会层面 · safety论文，上一届是 130 篇。

CVPR 2026 有多少篇社会层面 · safety论文？

CVPR 2026 有 3 篇社会层面 · safety论文，上一届是 1 篇。

哪个会议的社会层面 · safety论文最多？

ICML，2023–2026 年共 364 篇。