推理加速论文趋势 2023–2026

主要结论

• 推理加速论文从 2023 年的 54 篇增长到 2026 年的 91 篇，三年间扩大到约 1.7 倍。
• NeurIPS 是发推理加速论文最多的会议，2023–2026 年共 90 篇，占四个会议总量的 35%。
• 交叉方向里涨得最快的是深度学习 · large language models：同时属于两个主题的论文从 2023 年的 1 篇增至 2026 年的 9 篇。
• 在「效率与系统」的 5 个主题里，它的 2026 年论文数排第 5；同组增长最快的硬件协同是 4.4 倍，推理加速是 1.7 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 Method-Class Divide in Sub-4-Bit Quantization:\\Iterative vs.\ Single-Pass Sensitivity to Data Composition
• ICML 2026Consistency Deep Equilibrium Models
• ICLR 2026OrderDP: A Theoretically Guaranteed Lossless Dynamic Data Pruning Framework
• CVPR 2026GeoRK2 Geometry-Guided Runge-Kutta Integration for Diffusion Transformer Acceleration
• NeurIPS 2025PolarQuant: Leveraging Polar Transformation for Key Cache Quantization and Decoding Acceleration
• ICML 2025RePaViT: Scalable Vision Transformer Acceleration via Structural Reparameterization on Feedforward Network Layers
• ICLR 2025Progressive distillation induces an implicit curriculum
• CVPR 2025Adapting Dense Matching for Homography Estimation with Grid-based Acceleration
• NeurIPS 2024Provable Acceleration of Nesterov's Accelerated Gradient for Asymmetric Matrix Factorization and Linear Neural Networks
• ICML 2024Diversified Batch Selection for Training Acceleration
• ICLR 2024Embarrassingly Simple Dataset Distillation
• CVPR 2024Data-Free Quantization via Pseudo-label Filtering
• NeurIPS 2023Continuous-time Analysis of Anchor Acceleration
• ICML 2023Auxiliary Modality Learning with Generalized Curriculum Distillation
• ICLR 2023Accurate Neural Training with 4-bit Matrix Multiplications at Standard Formats
• CVPR 2023NIPQ: Noise Proxy-Based Integrated Pseudo-Quantization
• NeurIPS 2026AcceleGrad#: Adaptive Geometry-Aware Acceleration
• ICML 2026PsumQuant: In-line Post-training Partial Sum Quantizer for Energy Efficient NPU Inference
• ICLR 2026NLI : Non-uniform Linear Interpolation Approximation of Nonlinear Operations for Efficient LLMs Inference
• CVPR 2026SODA Sensitivity-Oriented Dynamic Acceleration for Diffusion Transformer

全部推理加速论文

共 17 篇，第 1 / 1 页

1. ICLR 2026BBQ: Boosting Quantization Entropy with Bell Box Quantization
2. ICLR 2026Corner Gradient Descent
3. ICLR 2026Dataset Distillation as Pushforward Optimal Quantization
4. ICLR 2026DiCache: Let Diffusion Model Determine Its Own Cache
5. ICLR 2026Diffusion LLMs Can Do Faster-Than-AR Inference via Discrete Diffusion Forcing
6. ICLR 2026Distributionally Robust Linear Regression with Block Lewis Weights
7. ICLR 2026Egalitarian Gradient Descent: A Simple Approach to Accelerated Grokking
8. ICLR 2026Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding
9. ICLR 2026GAS: Improving Discretization of Diffusion ODEs via Generalized Adversarial Solver
10. ICLR 2026NLI : Non-uniform Linear Interpolation Approximation of Nonlinear Operations for Efficient LLMs Inference
11. ICLR 2026OrderDP: A Theoretically Guaranteed Lossless Dynamic Data Pruning Framework
12. ICLR 2026RAPID$^3$: Tri-Level Reinforced Acceleration Policies for Diffusion Transformer
13. ICLR 2026Score Distillation Beyond Acceleration: Generative Modeling from Corrupted Data
14. ICLR 2026SGD with Adaptive Preconditioning: Unified Analysis and Momentum Acceleration
15. ICLR 2026Towards a Universally Transferable Acceleration Method for Density Functional Theory
16. ICLR 2026TurboQuant: Online Vector Quantization with Near-optimal Distortion Rate
17. ICLR 2026Universal Inverse Distillation for Matching Models with Real-Data Supervision (No GANs)

常见问题

2023–2026 年四大会议一共有多少篇推理加速论文？

共 259 篇，按年份分别是 2023 年 54 篇、2024 年 50 篇、2025 年 64 篇、2026 年 91 篇。

NeurIPS 2026 有多少篇推理加速论文？

NeurIPS 2026 有 33 篇推理加速论文，上一届是 15 篇。

ICLR 2026 有多少篇推理加速论文？

ICLR 2026 有 17 篇推理加速论文，上一届是 18 篇。

ICML 2026 有多少篇推理加速论文？

ICML 2026 有 26 篇推理加速论文，上一届是 22 篇。

CVPR 2026 有多少篇推理加速论文？

CVPR 2026 有 15 篇推理加速论文，上一届是 9 篇。

哪个会议的推理加速论文最多？

NeurIPS，2023–2026 年共 90 篇。