芯片设计 ML论文趋势 2023–2026

主要结论

• 芯片设计 ML论文从 2023 年的 28 篇增长到 2026 年的 64 篇，三年间扩大到约 2.3 倍。
• NeurIPS 是发芯片设计 ML论文最多的会议，2023–2026 年共 83 篇，占四个会议总量的 46%。
• 交叉方向里涨得最快的是扩散模型：同时属于两个主题的论文从 2023 年的 0 篇增至 2026 年的 9 篇。
• 在「硬件与编译」的 5 个主题里，它的 2026 年论文数排第 2；同组增长最快的端侧推理是 4.0 倍，芯片设计 ML是 2.3 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 2026ChiP-STAR: Spatial-Topological Attention for Pre-trained Generative Chip Routing
• ICML 2026Order Matters: Unveiling the Hidden Impact of Macro Placement Sequences via Proxy-Guided LLM Evolution
• ICLR 2026How Muon’s Spectral Design Benefits Generalization: A Study on Imbalanced Data
• CVPR 2026PSDesigner Automated Graphic Design with a Human-Like Creative Workflow
• NeurIPS 2025MIHC: Multi-View Interpretable Hypergraph Neural Networks with Information Bottleneck for Chip Congestion Prediction
• ICML 2025Probabilistic Factorial Experimental Design for Combinatorial Interventions
• ICLR 2025LaMPlace: Learning to Optimize Cross-Stage Metrics in Macro Placement
• CVPR 2025From Elements to Design: A Layered Approach for Automatic Graphic Design Composition
• NeurIPS 2024AFBench: A Large-scale Benchmark for Airfoil Design
• ICML 2024A Circuit Domain Generalization Framework for Efficient Logic Synthesis in Chip Design
• ICLR 2024CircuitNet 2.0: An Advanced Dataset for Promoting Machine Learning Innovations in Realistic Chip Design Environment
• CVPR 2024Discontinuity-preserving Normal Integration with Auxiliary Edges
• NeurIPS 2023Circuit as Set of Points
• ICML 2023AbODE: Ab initio antibody design using conjoined ODEs
• ICLR 2023Neural Design for Genetic Perturbation Experiments
• CVPR 2023Visibility Constrained Wide-Band Illumination Spectrum Design for Seeing-in-the-Dark
• NeurIPS 2026DiffCool: Label-Free Synthesis of Chip-Tailored Heat Sinks via Thermal-Aware Diffusion
• ICML 2026Axiomatic Atlas: A Prescriptive Framework for Neural Architecture Design
• ICLR 2026Topology Matters in RTL Circuit Representation Learning
• CVPR 2026FEAT Fashion Editing and Try-On from Any Design

全部芯片设计 ML论文

共 34 篇，第 1 / 1 页

1. NeurIPS 2026$R^2E$: A Role-driven Reward Evolutionary Framework for Automated Reward Function Design
2. NeurIPS 2026A Generalized Tikhonov Layer for Interpretable-by-design Graph Neural Networks
3. NeurIPS 2026An Equivariance Principle for Optimizer Design: Symmetry-Compatible Updates for Embeddings, LM Heads, and MoE Routers
4. NeurIPS 2026AutoManifold: Agentic Design of Data Visualisation Algorithms via Manifold Embedding
5. NeurIPS 2026ChiP-STAR: Spatial-Topological Attention for Pre-trained Generative Chip Routing
6. NeurIPS 2026Computational Dynamic Mechanism Design
7. NeurIPS 2026DiffCool: Label-Free Synthesis of Chip-Tailored Heat Sinks via Thermal-Aware Diffusion
8. NeurIPS 2026Differentiating Network Design Objectives for Balancing Cost and Distance
9. NeurIPS 2026EGCA: A Spectral Perspective on Forward Process Design in Diffusion Models
10. NeurIPS 2026GenAI Evaluation Results are Largely Artifacts of Evaluation Design Choices: Evidence from Audit Studies of Resume Screening
11. NeurIPS 2026HSCO-Bench: An Agent-Driven End-to-End Hardware-Software Co-design Benchmark for Systems-on-Chip
12. NeurIPS 2026Large Language Models Enhanced Covariate-adjusted Response-adaptive Randomization Design
13. NeurIPS 2026LLM Judge Validation Under Sparse Overlap: From Inference to Design
14. NeurIPS 2026Locality-Controlled OOD Guidance for Robust Regulatory DNA Sequence Design
15. NeurIPS 2026Maximin Robust Bayesian Experimental Design
16. NeurIPS 2026Mechanism Design for AI Overviews: Creator Incentives and Long-Term Profit
17. NeurIPS 2026Mechanistic Critics for Sample-Efficient NPU Design-Space Exploration
18. NeurIPS 2026Mitigating Overgeneralization in RND via Spectral Target Design
19. NeurIPS 2026Neural-Corrected Operator Learning for Homogenization and Inverse Design
20. NeurIPS 2026On the Meta-Design of Allocation Problems
21. NeurIPS 2026PCBInnoBench: Benchmarking LLM Agents on Real-World PCB Design
22. NeurIPS 2026Principled Design of Diffusion-based Optimizers for Inverse Problems
23. NeurIPS 2026Refactoring Code Through Library Design
24. NeurIPS 2026Regret Minimization in Single-Dimensional Contract-Design with Binary Actions
25. NeurIPS 2026Robust-by-Design Distributional Learning from Contaminated Samples
26. NeurIPS 2026Sample Size Design for Bounds on Discrete Probabilities of Causation
27. NeurIPS 2026Scattered by Design: Why Per-Output Pruning Resists Structured Compression
28. NeurIPS 2026Selectivity and Shape in the Design of Forward-Forward Goodness Functions
29. NeurIPS 2026SimpleEvol: Efficient Intelligence Conversion via Less Human Prior in Automated Heuristic Design
30. NeurIPS 2026The Alternation Depth Principle for Neural Operator Design
31. NeurIPS 2026The Design Space of Tri-Modal Masked Diffusion Models
32. NeurIPS 2026VEHBench: A Stage-Local Diagnostic Benchmark for LLM-Assisted Vibration Energy Harvester Design
33. NeurIPS 2026When Does Knowing the State Help? Diagnosing Process vs. Outcome Reward Design
34. NeurIPS 2026μLM: Rethinking Sub-100M Language Models through Memory-First Design

常见问题

2023–2026 年四大会议一共有多少篇芯片设计 ML论文？

共 179 篇，按年份分别是 2023 年 28 篇、2024 年 43 篇、2025 年 44 篇、2026 年 64 篇。

NeurIPS 2026 有多少篇芯片设计 ML论文？

NeurIPS 2026 有 34 篇芯片设计 ML论文，上一届是 20 篇。

ICLR 2026 有多少篇芯片设计 ML论文？

ICLR 2026 有 9 篇芯片设计 ML论文，上一届是 9 篇。

ICML 2026 有多少篇芯片设计 ML论文？

ICML 2026 有 18 篇芯片设计 ML论文，上一届是 9 篇。

CVPR 2026 有多少篇芯片设计 ML论文？

CVPR 2026 有 3 篇芯片设计 ML论文，上一届是 6 篇。

哪个会议的芯片设计 ML论文最多？

NeurIPS，2023–2026 年共 83 篇。