缩放定律论文趋势 2023–2026

主要结论

• 缩放定律论文从 2023 年的 60 篇增长到 2026 年的 248 篇，三年间扩大到约 4.1 倍。
• NeurIPS 是发缩放定律论文最多的会议，2023–2026 年共 221 篇，占四个会议总量的 37%。
• 交叉方向里涨得最快的是时间序列：同时属于两个主题的论文从 2023 年的 1 篇增至 2026 年的 24 篇。
• 在「优化与缩放」的 5 个主题里，它的 2026 年论文数排第 5；同组增长最快的低秩适配是 3.0 倍，缩放定律是 4.1 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 Unified Merge Calculus for Learning-Rate Scaling on Neural Computation Graphs
• ICML 2026A Capacity-Based Rationale for Multi-Head Attention
• ICLR 2026$\mathbf{Li_2}$: A Framework on Dynamics of Feature Emergence and Delayed Generalization
• CVPR 2026A Unified Framework for Knowledge Transfer in Bidirectional Model Scaling
• NeurIPS 2025Audits Under Resource, Data, and Access Constraints: Scaling Laws For Less Discriminatory Alternatives
• ICML 2025A Simple Model of Inference Scaling Laws
• ICLR 2025(Mis)Fitting Scaling Laws: A Survey of Scaling Law Fitting Techniques in Deep Learning
• CVPR 2025D^3: Scaling Up Deepfake Detection by Learning from Discrepancy
• NeurIPS 2024An exactly solvable model for emergence and scaling laws in the multitask sparse parity problem
• ICML 2024A Dynamical Model of Neural Scaling Laws
• ICLR 2024Principled Architecture-aware Scaling of Hyperparameters
• CVPR 2024From Activation to Initialization: Scaling Insights for Optimizing Neural Fields
• NeurIPS 2023Abide by the law and follow the flow: conservation laws for gradient flows
• ICML 2023Dissecting the Effects of SGD Noise in Distinct Regimes of Deep Learning
• ICLR 2023Distributionally Robust Post-hoc Classifiers under Prior Shifts
• CVPR 2023LargeKernel3D: Scaling Up Kernels in 3D Sparse CNNs
• NeurIPS 2026ADIS-Law: Unified Scaling Laws for Annealing-Phase Domain Injection in Large Language Models
• ICML 2026Balancing Learning Rates Across Layers: Exact Two-Step Dynamics and Optimal Scaling in Linear Neural Networks
• ICLR 2026A universal compression theory: Lottery ticket hypothesis and superpolynomial scaling laws
• CVPR 2026MatAnyone 2 Scaling Video Matting via a Learned Quality Evaluator

全部缩放定律论文

共 150 篇，第 2 / 3 页

51. ICLR 2026Scaling Laws of SignSGD in Linear Regression: When Does It Outperform SGD?
52. ICLR 2026Scaling Laws Revisited: Modeling the Role of Data Quality in Language Model Pretraining
53. ICLR 2026Scaling Linear Attention with Sparse State Expansion
54. ICLR 2026Scaling Reasoning Hop Exposes Weaknesses: Demystifying and Improving Hop Generalization in Large Language Models
55. ICLR 2026Scaling Sequence-to-Sequence Generative Neural Rendering
56. ICLR 2026Scaling with Collapse: Efficient and Predictable Training of LLM Families
57. ICLR 2026STEM: SCALING TRANSFORMERS WITH EMBEDDING MODULES
58. ICLR 2026The Hot Mess of AI: How Does Misalignment Scale With Model Intelligence and Task Complexity?
59. ICLR 2026The Limits of Inference Scaling Through Resampling
60. ICLR 2026The Serial Scaling Hypothesis
61. ICLR 2026Theory of Scaling Laws for In-Context Regression: Depth, Width, Context and Time
62. ICLR 2026Toward Principled Flexible Scaling for Self-Gated Neural Activation
63. ICLR 2026Towards Greater Leverage: Scaling Laws for Efficient Mixture-of-Experts Language Models
64. ICLR 2026TrimR: Verifier-based Training-Free Thinking Trimming for Efficient Test-Time Scaling
65. ICLR 2026UltraMemV2: Memory Networks Scaling to 120B Parameters with Superior Long-Context Learning
66. ICLR 2026Understanding the Mechanisms of Fast Hyperparameter Transfer
67. ICLR 2026Unveiling Downstream Performance Scaling of LLMs: A Clustering-Based Perspective
68. ICLR 2026UrbanVerse: Scaling Urban Simulation by Watching City-Tour Videos
69. ICLR 2026WATS: Wavelet-Aware Temperature Scaling for Reliable Graph Neural Networks
70. ICLR 2026Webscale-RL: Automated Data Pipeline for Scaling RL Data to Pretraining Levels
71. ICLR 2026What Scales in Cross-Entropy Scaling Law?
72. ICLR 2026Why Less is More (Sometimes): A Theory of Data Curation
73. ICLR 2026xLSTM Scaling Laws: Competitive Performance with Linear Time-Complexity
74. ICLR 2025(Mis)Fitting Scaling Laws: A Survey of Scaling Law Fitting Techniques in Deep Learning
75. ICLR 2025A Multi-Power Law for Loss Curve Prediction Across Learning Rate Schedules
76. ICLR 2025A Solvable Attention for Neural Scaling Laws
77. ICLR 2025Adaptive Data Optimization: Dynamic Sample Selection with Scaling Laws
78. ICLR 2025Analyzing Neural Scaling Laws in Two-Layer Networks with Power-Law Data Spectra
79. ICLR 2025Beyond Model Collapse: Scaling Up with Synthesized Data Requires Verification
80. ICLR 2025Breaking Neural Network Scaling Laws with Modularity
81. ICLR 2025Compute-Constrained Data Selection
82. ICLR 2025Compute-Optimal LLMs Provably Generalize Better with Scale
83. ICLR 2025Data Mixing Laws: Optimizing Data Mixtures by Predicting Language Modeling Performance
84. ICLR 2025Data Scaling Laws in Imitation Learning for Robotic Manipulation
85. ICLR 2025DCT-CryptoNets: Scaling Private Inference in the Frequency Domain
86. ICLR 2025Fast Training of Sinusoidal Neural Fields via Scaling Initialization
87. ICLR 2025GETS: Ensemble Temperature Scaling for Calibration in Graph Neural Networks
88. ICLR 2025High-dimensional Analysis of Knowledge Distillation: Weak-to-Strong Generalization and Scaling Laws
89. ICLR 2025How Feature Learning Can Improve Neural Scaling Laws
90. ICLR 2025How Much is a Noisy Image Worth? Data Scaling Laws for Ambient Diffusion.
91. ICLR 2025Inference Optimal VLMs Need Fewer Visual Tokens and More Parameters
92. ICLR 2025Inference Scaling Laws: An Empirical Analysis of Compute-Optimal Inference for LLM Problem-Solving
93. ICLR 2025Language models scale reliably with over-training and on downstream tasks
94. ICLR 2025LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging
95. ICLR 2025On Scaling Up 3D Gaussian Splatting Training
96. ICLR 2025OSTQuant: Refining Large Language Model Quantization with Orthogonal and Scaling Transformations for Better Distribution Fitting
97. ICLR 2025Physics of Language Models: Part 3.3, Knowledge Capacity Scaling Laws
98. ICLR 2025Progressive Token Length Scaling in Transformer Encoders for Efficient Universal Segmentation
99. ICLR 2025Quest: Query-centric Data Synthesis Approach for Long-context Scaling of Large Language Model
100. ICLR 2025Reasoning with Latent Thoughts: On the Power of Looped Transformers

常见问题

2023–2026 年四大会议一共有多少篇缩放定律论文？

共 605 篇，按年份分别是 2023 年 60 篇、2024 年 96 篇、2025 年 201 篇、2026 年 248 篇。

NeurIPS 2026 有多少篇缩放定律论文？

NeurIPS 2026 有 82 篇缩放定律论文，上一届是 80 篇。

ICLR 2026 有多少篇缩放定律论文？

ICLR 2026 有 73 篇缩放定律论文，上一届是 48 篇。

ICML 2026 有多少篇缩放定律论文？

ICML 2026 有 78 篇缩放定律论文，上一届是 56 篇。

CVPR 2026 有多少篇缩放定律论文？

CVPR 2026 有 15 篇缩放定律论文，上一届是 17 篇。

哪个会议的缩放定律论文最多？

NeurIPS，2023–2026 年共 221 篇。