深度学习 · probabilistic methods论文趋势 2023–2026

主要结论

• 深度学习 · probabilistic methods论文从 2023 年的 107 篇增长到 2026 年的 107 篇，三年间扩大到约 1.0 倍。
• NeurIPS 是发深度学习 · probabilistic methods论文最多的会议，2023–2026 年共 440 篇，占四个会议总量的 69%。
• 交叉方向里涨得最快的是驾驶仿真：同时属于两个主题的论文从 2023 年的 5 篇增至 2026 年的 14 篇。
• 在「深度学习」的 11 个主题里，它的 2026 年论文数排第 9；同组增长最快的扩散模型是 4.1 倍，深度学习 · probabilistic methods是 1.0 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 Faster Algorithm for the Half-Trek Criterion in Structural Causal Models
• ICML 2026An Efficient Joint Learning Approach for Item Response Theory
• ICLR 2026Probabilistic Kernel Function for Fast Angle Testing
• CVPR 2026Beyond Text Visual Description Assembly by Probabilistic Model for CLIP-based
• NeurIPS 2025$\text{G}^2\text{M}$: A Generalized Gaussian Mirror Method to Boost Feature Selection Power
• ICML 2025Adjusting Model Size in Continual Gaussian Processes: How Big is Big Enough?
• ICLR 2025P-SPIKESSM: HARNESSING PROBABILISTIC SPIKING STATE SPACE MODELS FOR LONG-RANGE DEPENDENCY TASKS
• CVPR 2025ProbPose: A Probabilistic Approach to 2D Human Pose Estimation
• NeurIPS 2024[Re] GNNInterpreter: A probabilistic generative model-level explanation for Graph Neural Networks
• ICML 2024A Probabilistic Approach to Learning the Degree of Equivariance in Steerable CNNs
• ICLR 2024A Probabilistic Framework for Modular Continual Learning
• CVPR 2024CAPE: CAM as a Probabilistic Ensemble for Enhanced DNN Interpretation
• NeurIPS 2023A Computationally Efficient Sparsified Online Newton Method
• ICML 2023A theory of representation learning gives a deep generalisation of kernel methods
• ICLR 2023A Laplace-inspired Distribution on SO(3) for Probabilistic Rotation Estimation
• CVPR 2023Probabilistic Knowledge Distillation of Face Ensembles
• NeurIPS 2026Affine Tracing: A New Paradigm for Probabilistic Linear Solvers
• ICML 2026Avoid What You Know: Divergent Trajectory Balance for GFlowNets
• ICLR 2026Learning Survival Distributions with Individually Calibrated Asymmetric Laplace Distribution
• CVPR 2026ARES Unifying Asymmetric RGB-Event Stereo for Probabilistic Scene Flow Estimation

全部深度学习 · probabilistic methods论文

共 107 篇，第 2 / 3 页

51. ICML 2026Learning to Bet for Horizon-Aware Anytime-Valid Testing
52. ICML 2026LieStoNet: Learning Lie Symmetries from Spatiotemporal Data for Stochastic Dynamical Systems
53. ICML 2026Maximum-Likelihood Learning of Latent Dynamics Without Reconstruction
54. ICML 2026Minimum Distance Summaries for Robust Neural Posterior Estimation
55. ICML 2026Monotonic Variational Gaussian Process for Efficient Data Collection
56. ICML 2026Multi-Distribution Robust Conformal Prediction
57. ICML 2026Multiple Choice Learning of Low-Rank Adapters for Language Modeling
58. ICML 2026Near-Universal Multiplicative Updates for Nonnegative Einsum Factorization
59. ICML 2026On the Epistemic Uncertainty of Overparametrized Neural Networks
60. ICML 2026Parametric Prior Mapping Framework for Non-stationary Probabilistic Time Series Forecasting
61. ICML 2026PPDL: LLM-Based Flows as Probabilistic Programs
62. ICML 2026Predictable Compression Failures: Order Sensitivity and Information Budgeting for Evidence-Grounded Binary Adjudication
63. ICML 2026Principled Confidence Estimation for Deep Computed Tomography
64. ICML 2026Probabilistic Bisection Algorithm Provably Achieves Exponential Convergence
65. ICML 2026Probabilistic Retrofitting of Learned Simulators
66. ICML 2026Probably Approximately Correct Labels
67. ICML 2026ProbMoE: Differentiable Probabilistic Routing for Mixture-of-Experts
68. ICML 2026Pseudo-Mallows for Efficient Probabilistic Preference Learning
69. ICML 2026Reduction of Probabilistic Chemical Reaction Networks
70. ICML 2026Revisiting Neural Processes via Fourier Transform and Volterra Series
71. ICML 2026Robust Bayes-Assisted Conformal Prediction
72. ICML 2026Set-Preserving Calibration from Conformal P-Values to E-Values
73. ICML 2026Simultaneous Confidence Bounds for Aggregated Effects via Exact Subset Optimization
74. ICML 2026Spectral Flow Matching: Stabilizing Stochastic GFlowNets via Frequency-Domain Regularization
75. ICML 2026SpeedCP: Fast Kernel-based Conditional Conformal Prediction
76. ICML 2026Split Group Knockoffs: Controlling False Discovery Rate in Transformational Group Sparsity
77. ICML 2026SVRG and Beyond via Posterior Correction
78. ICML 2026Tensor Train Diffusion: Leveraging Low-Rank Structures for High-Dimensional Score-Based Sampling
79. ICML 2026Transformers Can Learn Posterior Predictive Distributions In-Context
80. NeurIPS 2026A Faster Algorithm for the Half-Trek Criterion in Structural Causal Models
81. NeurIPS 2026A probabilistic model of visual segmentation explains early visual cortical dynamics
82. NeurIPS 2026Affine Tracing: A New Paradigm for Probabilistic Linear Solvers
83. NeurIPS 2026Announced Breaks Separate Conformal Reliability from Frequency Calibration
84. NeurIPS 2026Auditing Conformal Prediction for Language Models Under Inference-Time Model Shift
85. NeurIPS 2026CoTs as Probabilistic Programs: A Programmatic View of Thinking Step-by-Step in Language Models
86. NeurIPS 2026Efficient Adaptive Data Acquisition via Pretrained Belief Representations
87. NeurIPS 2026Exact Gaussian Moment Matching for Residual Networks: a Second-Order Method
88. NeurIPS 2026Filtered-Trace Online Variational Training for Probabilistic Spiking Neural Networks
89. NeurIPS 2026FlowBatt: Flow Matching for Probabilistic Battery Degradation Prediction
90. NeurIPS 2026Heterogeneous Judge-Aware Ranking with Sensitivity, Disagreement, and Confidence
91. NeurIPS 2026Impossibility of Distribution-Free Predictive Inference for Individual Treatment Effects
92. NeurIPS 2026Inference and estimation with unidentifiable latent treatment effects
93. NeurIPS 2026Just Ramp-Up: Debiasing Regression-based Estimator for A/B Tests under Network Interference
94. NeurIPS 2026One-Shot Generative Flows: Existence and Obstructions
95. NeurIPS 2026Probabilistic Recursive Reasoning
96. NeurIPS 2026Probabilistic Signature Inversion: Learning Conditional Distributions from Truncated Signatures
97. NeurIPS 2026Probabilistic Tiny Recursive Model
98. NeurIPS 2026probly: Uncertainty-Aware Machine Learning
99. NeurIPS 2026Robust Conditional Conformal Prediction via Branched Normalizing Flow
100. NeurIPS 2026Scalable Variational Bayesian Fine-Tuning of LLMs via Orthogonalized Low-Rank Adapters

常见问题

2023–2026 年四大会议一共有多少篇深度学习 · probabilistic methods论文？

共 637 篇，按年份分别是 2023 年 107 篇、2024 年 160 篇、2025 年 263 篇、2026 年 107 篇。

NeurIPS 2026 有多少篇深度学习 · probabilistic methods论文？

NeurIPS 2026 有 28 篇深度学习 · probabilistic methods论文，上一届是 212 篇。

ICLR 2026 有多少篇深度学习 · probabilistic methods论文？

ICLR 2026 有 13 篇深度学习 · probabilistic methods论文，上一届是 10 篇。

ICML 2026 有多少篇深度学习 · probabilistic methods论文？

ICML 2026 有 59 篇深度学习 · probabilistic methods论文，上一届是 37 篇。

CVPR 2026 有多少篇深度学习 · probabilistic methods论文？

CVPR 2026 有 7 篇深度学习 · probabilistic methods论文，上一届是 4 篇。

哪个会议的深度学习 · probabilistic methods论文最多？

NeurIPS，2023–2026 年共 440 篇。