可微推理论文趋势 2023–2026

主要结论

• 可微推理论文从 2023 年的 31 篇增长到 2026 年的 79 篇，三年间扩大到约 2.5 倍。
• NeurIPS 是发可微推理论文最多的会议，2023–2026 年共 65 篇，占四个会议总量的 35%。
• 交叉方向里涨得最快的是数学推理：同时属于两个主题的论文从 2023 年的 3 篇增至 2026 年的 29 篇。
• 在「神经符号」的 5 个主题里，它的 2026 年论文数排第 3；同组增长最快的数学推理是 10 倍，可微推理是 2.5 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 2026Beyond MSE: Differentiable Complexity Priors for Structure-Preserving Neural Denoising
• ICML 2026Differentiable Conformal Training for LLM Reasoning Factuality
• ICLR 2026$\boldsymbol{\partial^\infty}$-Grid: A Neural Differential Equation Solver with Differentiable Feature Grids
• CVPR 2026D-Convexity A Unified Differentiable Convex Shape Prior via Quasi-Concavity for
• NeurIPS 2025Consistency Conditions for Differentiable Surrogate Losses
• ICML 2025DiLQR: Differentiable Iterative Linear Quadratic Regulator via Implicit Differentiation
• ICLR 2025Analytic DAG Constraints for Differentiable DAG Learning
• CVPR 2025Differentiable Inverse Rendering with Interpretable Basis BRDFs
• NeurIPS 2024DeiSAM: Segment Anything with Deictic Prompting
• ICML 2024A Differentiable Partially Observable Generalized Linear Model with Forward-Backward Message Passing
• ICLR 2024Generalized Neural Sorting Networks with Error-Free Differentiable Swap Functions
• CVPR 2024Differentiable Display Photometric Stereo
• NeurIPS 2023Differentiable Clustering with Perturbed Spanning Forests
• ICML 2023End-to-end Differentiable Clustering with Associative Memories
• ICLR 2023Learning Group Importance using the Differentiable Hypergeometric Distribution
• CVPR 2023Deep Learning of Partial Graph Matching via Differentiable Top-K
• NeurIPS 2026Differentiable Exact Learning of Algorithms
• ICML 2026Hide&Seek: Learning to Explain in an End-to-End Differentiable Network
• ICLR 2026Tversky Neural Networks: Psychologically Plausible Deep Learning with Differentiable Tversky Similarity
• CVPR 2026DiffBMP Differentiable Rendering with Bitmap Primitives

全部可微推理论文

共 184 篇，第 2 / 4 页

51. NeurIPS 2026Anchoring Reasoning Distillation via Syntactic Constraints
52. NeurIPS 2026Beyond MSE: Differentiable Complexity Priors for Structure-Preserving Neural Denoising
53. NeurIPS 2026Block-Wise Differentiable Sinkhorn Attention: Tail-Refinement Gradients with a Gap-Aware Dustbin Bridge
54. NeurIPS 2026Can Agents Price a Reaction? Evaluating LLMs on Chemical Cost Reasoning
55. NeurIPS 2026Can Linguistic Reasoning Vectors Enhance Multimodal Reasoning Ability?
56. NeurIPS 2026Commit Then Explore: Reasoning Models Diversify, Not Converge
57. NeurIPS 2026Convex Compositional Reasoning Models
58. NeurIPS 2026Credit Assignment with Resets in Language Model Reasoning
59. NeurIPS 2026Differentiable Belief-based Opponent Shaping
60. NeurIPS 2026Differentiable Exact Learning of Algorithms
61. NeurIPS 2026Differentiable Nonlinear Model Predictive Control
62. NeurIPS 2026Differentiable Range-Partition Entropy for Entropy-Sensitive Geometric Algorithms
63. NeurIPS 2026Diving-R1: Empowering Multimodal LLMs with Traceable Progressive Reasoning for Interpretable Diving Action Quality Assessment
64. NeurIPS 2026From Outcome to Representation: Tracing Reasoning Mechanisms through Integrated Policy Gradient
65. NeurIPS 2026Iterative Nonlinear Computation Underlying Abstract Reasoning
66. NeurIPS 2026Latent Spatial Reasoning: Building Innate 3D Awareness via Latent-Space Distillation
67. NeurIPS 2026LEVDA: Latent Ensemble Variational Data Assimilation via Differentiable Dynamics
68. NeurIPS 2026Mitigating Overthinking in Large Reasoning Language Models via Reasoning Path Deviation Monitoring
69. NeurIPS 2026Natural Synthesis: Outperforming Reactive Synthesis Tools with Large Reasoning Models
70. NeurIPS 2026Neurosymbolic Object-Centric Learning with Distant Supervision
71. NeurIPS 2026On the Token Value Inequality in Efficient Reasoning
72. NeurIPS 2026OS-Pruner: Pruning Chains-of-Thought of Reasoning Models via Optimal Stopping
73. NeurIPS 2026Quantized Reasoning Models Think They Need to Think Longer, but They Do Not
74. NeurIPS 2026Reasoning with Undecoded Tokens in Diffusion Language Models
75. NeurIPS 2026SphMind: Towards Robust, Training-Free VLM-based Spatial Reasoning with a 360 Camera
76. NeurIPS 2026State of Thought Enables Endogenous Reasoning
77. NeurIPS 2026Strategist: Designing Agentic Reasoning at Scale
78. NeurIPS 2026VI-CuRL: Stabilizing Verifier-Independent RL Reasoning via Confidence-Guided Variance Reduction
79. CVPR 2025Differentiable Inverse Rendering with Interpretable Basis BRDFs
80. CVPR 2025Explainable Saliency: Articulating Reasoning with Contextual Prioritization
81. CVPR 2025LayoutVLM: Differentiable Optimization of 3D Layout via Vision-Language Models
82. CVPR 2025PyTorchGeoNodes: Enabling Differentiable Shape Programs for 3D Shape Reconstruction
83. CVPR 2025Reasoning to Attend: Try to Understand How <SEG> Token Works
84. CVPR 2025Shape Abstraction via Marching Differentiable Support Functions
85. ICLR 2025Analytic DAG Constraints for Differentiable DAG Learning
86. ICLR 2025BlendRL: A Framework for Merging Symbolic and Neural Policy Learning
87. ICLR 2025Chemistry-Inspired Diffusion with Non-Differentiable Guidance
88. ICLR 2025Clique Number Estimation via Differentiable Functions of Adjacency Matrix Permutations
89. ICLR 2025Differentiable Causal Discovery for Latent Hierarchical Causal Models
90. ICLR 2025Differentiable Integer Linear Programming
91. ICLR 2025Differentiable Rule Induction from Raw Sequence Inputs
92. ICLR 2025DPaI: Differentiable Pruning at Initialization with Node-Path Balance Principle
93. ICLR 2025Generalizing Reasoning Problems to Longer Lengths
94. ICLR 2025Large Language Models Meet Symbolic Provers for Logical Reasoning Evaluation
95. ICLR 2025Linear Mode Connectivity in Differentiable Tree Ensembles
96. ICLR 2025PRDP: Progressively Refined Differentiable Physics
97. ICLR 2025ReMoE: Fully Differentiable Mixture-of-Experts with ReLU Routing
98. ICML 2025Can Transformers Reason Logically? A Study in SAT Solving
99. ICML 2025Deliberation in Latent Space via Differentiable Cache Augmentation
100. ICML 2025Differentiable Structure Learning with Ancestral Constraints

常见问题

2023–2026 年四大会议一共有多少篇可微推理论文？

共 185 篇，按年份分别是 2023 年 31 篇、2024 年 29 篇、2025 年 46 篇、2026 年 79 篇。

NeurIPS 2026 有多少篇可微推理论文？

NeurIPS 2026 有 28 篇可微推理论文，上一届是 22 篇。

ICLR 2026 有多少篇可微推理论文？

ICLR 2026 有 16 篇可微推理论文，上一届是 13 篇。

ICML 2026 有多少篇可微推理论文？

ICML 2026 有 14 篇可微推理论文，上一届是 5 篇。

CVPR 2026 有多少篇可微推理论文？

CVPR 2026 有 21 篇可微推理论文，上一届是 6 篇。

哪个会议的可微推理论文最多？

NeurIPS，2023–2026 年共 65 篇。