社会层面 · accountability transparency and interpretability论文趋势 2023–2026

主要结论

• 社会层面 · accountability transparency and interpretability论文从 2023 年的 301 篇增长到 2026 年的 1,357 篇，三年间扩大到约 4.5 倍。
• NeurIPS 是发社会层面 · accountability transparency and interpretability论文最多的会议，2023–2026 年共 1,275 篇，占四个会议总量的 37%。
• 交叉方向里涨得最快的是社会层面 · privacy：同时属于两个主题的论文从 2023 年的 0 篇增至 2026 年的 93 篇。
• 在「对齐与负责任 AI」的 5 个主题里，它的 2026 年论文数排第 2；同组增长最快的社会层面 · safety是 93 倍，社会层面 · accountability transparency and interpretability是 4.5 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 2026A Unifying Perspective on Language Model Interpretability
• ICML 2026$\texttt{PRISM}$:A 3D Probabilistic Neural Representation for Interpretable Shape Modeling
• ICLR 2026Automated Interpretability Metrics Do Not Distinguish Trained and Random Transformers
• CVPR 2026Beyond Matching to Tiles Bridging Unaligned Aerial and Satellite Views
• NeurIPS 2025$\mathcal{X}^2$-DFD: A framework for e$\mathcal{X}$plainable and e$\mathcal{X}$tendable Deepfake Detection
• ICML 2025(How) Can Transformers Predict Pseudo-Random Numbers?
• ICLR 2025Adaptive Transformer Programs: Bridging the Gap Between Performance and Interpretability in Transformers
• CVPR 2025Attribute-formed Class-specific Concept Space: Endowing Language Bottleneck Model with Better Interpretability and Scalability
• NeurIPS 20242D-OOB: Attributing Data Contribution Through Joint Valuation Framework
• ICML 2024Attention Meets Post-hoc Interpretability: A Mathematical Perspective
• ICLR 2024"What Data Benefits My Classifier?" Enhancing Model Performance and Interpretability through Influence-Based Data Selection
• 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 Modern Look at the Relationship between Sharpness and Generalization
• ICLR 2023Deep Learning From Crowdsourced Labels: Coupled Cross-Entropy Minimization, Identifiability, and Regularization
• CVPR 2023Analyzing and Diagnosing Pose Estimation With Attributions
• NeurIPS 2026AdKnob: Ad Intensity Control and Labeling for LLM-Native Advertising
• ICML 2026$\texttt{ShaplEIG}$: Bayesian Experimental Design for Shapley Value Estimation
• ICLR 2026Bandits with Single-Peaked Preferences and Limited Resources
• CVPR 2026IEBGLAn Interpretability-Enhanced Brain Graph Learning Framework with LLM-Instructed Topology and

全部社会层面 · accountability transparency and interpretability论文

共 3,411 篇，第 19 / 69 页

901. ICML 2026Towards Steering without Sacrifice: Principled Training of Steering Vectors for Prompt-only Interventions
902. ICML 2026Towards the Explainability of Temporal Graph Networks via Memory Backtracking and Topological Attribution
903. ICML 2026Towards Whole-corpus Reconstruction of Heterogeneous RAG Knowledge Bases
904. ICML 2026TRACER: Trajectory Risk Aggregation for Critical Episodes in Agentic Reasoning
905. ICML 2026TraceRouter: Robust Safety for Large Foundation Models via Path-Level Intervention
906. ICML 2026Tracing the Persona Circuit: How Large Language Models Encode and Express Character Traits
907. ICML 2026Transformers learn factored representations
908. ICML 2026Transformers Provably Learn Algorithmic Solutions for Graph Connectivity, But Only with the Right Data
909. ICML 2026TSFAdv: Frequency-Guided Black-Box Adversarial Attacks on Time Series Forecasting
910. ICML 2026TT-Sparse: Learning Sparse Rule Models with Differentiable Truth Tables
911. ICML 2026TuneAhead: Predicting Fine-tuning Performance Before Training Begins
912. ICML 2026Turning Bias into Bugs: Bandit-Guided Style Manipulation Attacks on LLM Judges
913. ICML 2026Unified Time Series Explanations via Amortized Optimization and Instance-level Multi-Expert Knowledge Distillation
914. ICML 2026Universal Redundancies in Time Series Foundation Models
915. ICML 2026Unlearning Isn't Deletion: Investigating Reversibility of Machine Unlearning in LLMs
916. ICML 2026Unlearning Isn’t Forgetting: Revealing Hidden Leakage in Class Unlearning Evaluations
917. ICML 2026Unlearning’s Blind Spots: Over‑Unlearning and Prototypical Relearning Attack
918. ICML 2026VALUEFLOW: Toward Pluralistic and Steerable Value-based Alignment in Large Language Models
919. ICML 2026Variance-Reduced $(\varepsilon, \delta)-$Unlearning using Forget Set Gradients
920. ICML 2026Verified SHAP: Provable Bounds for Exact Shapley Values of Neural Networks
921. ICML 2026Visual Persuasion: What Influences Decisions of Vision-Language Models?
922. ICML 2026VPD-100K: Towards Generalizable and Fine-grained Visual Privacy Protection
923. ICML 2026WatchLog: Efficient and Interpretable Event Reasoning for Endpoint Detection and Response Logs with Multimodal LLMs
924. ICML 2026Watermarking Graph Neural Networks via Explanations for Ownership Protection
925. ICML 2026Watermarking LLM Agent Trajectories
926. ICML 2026Weight-sparse transformers have interpretable circuits
927. ICML 2026Weights to Code: Extracting Interpretable Algorithms from the Discrete Transformer
928. ICML 2026What is Missing? Explaining Neurons Activated by Absent Concepts
929. ICML 2026When Actions Go Off-Task: Detecting and Correcting Misaligned Actions in Computer-Use Agents
930. ICML 2026When and How Human Curation Backfires: Preference Alignment under Multi-Model Self-Consuming Loop
931. ICML 2026When Planning Fails Despite Correct Execution: On Epistemic Calibration for LLM-Based Multi-Agent Systems
932. ICML 2026When Random Saliency Looks Trained: Architectural Center Bias in CNN Interpretability
933. ICML 2026Who Evaluates AI's Social Impacts? Mapping Coverage and Gaps in First and Third Party Evaluations
934. ICML 2026Who Gets Credit or Blame? Attributing Accountability in Modern AI Systems
935. ICML 2026Who Transfers Safety? Identifying and Targeting Cross-Lingual Shared Safety Neurons
936. ICML 2026Who’s in Charge? Disempowerment Patterns in Real-World LLM Usage
937. ICML 2026Why Self-Distillation Helps and Hurts: Denoising vs. Signal Forgetting
938. ICML 2026WMVLM: Evaluating Diffusion Model Image Watermarking via Vision-Language Models
939. ICML 2026X-EviProbe: Post-hoc Parameter-Free Evidential Uncertainty Quantification for Frozen Graph Neural Networks
940. ICML 2026Z-Erase: Enabling Concept Erasure in Single Stream Diffusion Transformers
941. ICML 2026ZeroUnlearn: Few-Shot Knowledge Unlearning in Large Language Models
942. NeurIPS 2026$\mathcal{P}$Torch: Narrowing the Gap Between Projection and Gradient-Based Learning
943. NeurIPS 2026A Critical $\beta$-Scale for Posterior Collapse in Dirichlet $\beta$-VAEs
944. NeurIPS 2026A Mechanistic Interpretability Study of an Astronomical Foundation Model
945. NeurIPS 2026A simple model of co-emergence of grid and place fields
946. NeurIPS 2026A Unifying Perspective on Language Model Interpretability
947. NeurIPS 2026Achieving Optimal Static and Dynamic Regret Simultaneously in Bandits with Deterministic Losses
948. NeurIPS 2026Adapting to Conflict: Equilibrium Structure and Adaptive Learning in Harmonic Games
949. NeurIPS 2026AdKnob: Ad Intensity Control and Labeling for LLM-Native Advertising
950. NeurIPS 2026Agentic AI Design Should be Mediated to Promote Social Welfare

常见问题

2023–2026 年四大会议一共有多少篇社会层面 · accountability transparency and interpretability论文？

共 3,411 篇，按年份分别是 2023 年 301 篇、2024 年 589 篇、2025 年 1,164 篇、2026 年 1,357 篇。

NeurIPS 2026 有多少篇社会层面 · accountability transparency and interpretability论文？

NeurIPS 2026 有 416 篇社会层面 · accountability transparency and interpretability论文，上一届是 522 篇。

ICLR 2026 有多少篇社会层面 · accountability transparency and interpretability论文？

ICLR 2026 有 292 篇社会层面 · accountability transparency and interpretability论文，上一届是 211 篇。

ICML 2026 有多少篇社会层面 · accountability transparency and interpretability论文？

ICML 2026 有 541 篇社会层面 · accountability transparency and interpretability论文，上一届是 337 篇。

CVPR 2026 有多少篇社会层面 · accountability transparency and interpretability论文？

CVPR 2026 有 108 篇社会层面 · accountability transparency and interpretability论文，上一届是 94 篇。

哪个会议的社会层面 · accountability transparency and interpretability论文最多？

NeurIPS，2023–2026 年共 1,275 篇。