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 3 / 4 页

101. ICML 2025DiLQR: Differentiable Iterative Linear Quadratic Regulator via Implicit Differentiation
102. ICML 2025DOLPHIN: A Programmable Framework for Scalable Neurosymbolic Learning
103. NeurIPS 2025Can MLLMs Absorb Math Reasoning Abilities from LLMs as Free Lunch?
104. NeurIPS 2025Consistency Conditions for Differentiable Surrogate Losses
105. NeurIPS 2025Continuous Thought Machines
106. NeurIPS 2025Curriculum Abductive Learning
107. NeurIPS 2025Differentiable Decision Tree via "ReLU+Argmin" Reformulation
108. NeurIPS 2025Differentiable Generalized Sliced Wasserstein Plans
109. NeurIPS 2025Evaluating Program Semantics Reasoning with Type Inference in System $F$
110. NeurIPS 2025HypoBootstrap: A Bootstrapping Framework for Inductive Reasoning
111. NeurIPS 2025Improving Formal Reasoning of Transformer with State Stack
112. NeurIPS 2025LinPrim: Linear Primitives for Differentiable Volumetric Rendering
113. NeurIPS 2025Measuring the Faithfulness of Thinking Drafts in Large Reasoning Models
114. NeurIPS 2025Mind the Gap: Removing the Discretization Gap in Differentiable Logic Gate Networks
115. NeurIPS 2025MIRAGE: Assessing Hallucination in Multimodal Reasoning Chains of MLLM
116. NeurIPS 2025More Thinking, Less Seeing? Assessing Amplified Hallucination in Multimodal Reasoning Models
117. NeurIPS 2025MuSLR: Multimodal Symbolic Logical Reasoning
118. NeurIPS 2025Optimal Minimum Width for the Universal Approximation of Continuously Differentiable Functions by Deep Narrow MLPs
119. NeurIPS 2025Personalized Subgraph Federated Learning with Differentiable Auxiliary Projections
120. NeurIPS 2025Reasoning Models Better Express Their Confidence
121. NeurIPS 2025Reward Reasoning Models
122. NeurIPS 2025Right for the Right Reasons: Avoiding Reasoning Shortcuts via Prototypical Neurosymbolic AI
123. NeurIPS 2025Training Language Models to Reason Efficiently
124. NeurIPS 2025VeriThinker: Learning to Verify Makes Reasoning Model Efficient
125. CVPR 2024Cinematic Behavior Transfer via NeRF-based Differentiable Filming
126. CVPR 2024Differentiable Display Photometric Stereo
127. CVPR 2024Differentiable Micro-Mesh Construction
128. CVPR 2024Differentiable Point-based Inverse Rendering
129. CVPR 2024Dr. Bokeh: DiffeRentiable Occlusion-aware Bokeh Rendering
130. CVPR 2024Dr.Hair: Reconstructing Scalp-Connected Hair Strands without Pre-Training via Differentiable Rendering of Line Segments
131. CVPR 2024DUDF: Differentiable Unsigned Distance Fields with Hyperbolic Scaling
132. CVPR 2024Intraoperative 2D/3D Image Registration via Differentiable X-ray Rendering
133. CVPR 2024PixelLM: Pixel Reasoning with Large Multimodal Model
134. CVPR 2024TexTile: A Differentiable Metric for Texture Tileability
135. ICLR 2024COLEP: Certifiably Robust Learning-Reasoning Conformal Prediction via Probabilistic Circuits
136. ICLR 2024Differentiable Euler Characteristic Transforms for Shape Classification
137. ICLR 2024EBMDock: Neural Probabilistic Protein-Protein Docking via a Differentiable Energy Model
138. ICLR 2024Faithful Rule Extraction for Differentiable Rule Learning Models
139. ICLR 2024Generalized Neural Sorting Networks with Error-Free Differentiable Swap Functions
140. ICLR 2024Less is More: One-shot Subgraph Reasoning on Large-scale Knowledge Graphs
141. ICLR 2024Logical Languages Accepted by Transformer Encoders with Hard Attention
142. ICLR 2024Phenomenal Yet Puzzling: Testing Inductive Reasoning Capabilities of Language Models with Hypothesis Refinement
143. ICML 2024A Differentiable Partially Observable Generalized Linear Model with Forward-Backward Message Passing
144. ICML 2024Differentiable Combinatorial Scheduling at Scale
145. ICML 2024Differentiable Model Scaling using Differentiable Topk
146. ICML 2024Differentiable Weightless Neural Networks
147. ICML 2024Improving Neural Logic Machines via Failure Reflection
148. ICML 2024Scale-Free Image Keypoints Using Differentiable Persistent Homology
149. NeurIPS 2024Abductive Reasoning in Logical Credal Networks
150. NeurIPS 2024Convolutional Differentiable Logic Gate Networks

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。