代码生成论文趋势 2023–2026

程序合成与代码 LLM

统计范围：NeurIPS、ICLR、ICML、CVPR 四个会议，2023–2026 年，共 2,161 篇被归入该主题的论文。

主要结论

每年论文数

各会议论文数

按年份和会议统计

代码生成论文数，按年份和会议
年份NeurIPSICLRICMLCVPR合计
202329553195210
202413566111134446
202523016479172645
2026236241174209860

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

全部代码生成论文

年份全部2023202420252026
会议全部NeurIPSICLRICMLCVPR

共 526 篇，第 4 / 11 页

  1. ICLR 2026PCB-Bench: Benchmarking LLMs for Printed Circuit Board Placement and Routing
  2. ICLR 2026pi-Flow: Policy-Based Few-Step Generation via Imitation Distillation
  3. ICLR 2026PLAGUE: Plug-and-play Framework for Lifelong Adaptive Generation of Multi-turn Exploits
  4. ICLR 2026Plan then Act: Bi-level CAD Command Sequence Generation
  5. ICLR 2026Planned Diffusion
  6. ICLR 2026Planner Aware Path Learning in Diffusion Language Models Training
  7. ICLR 2026POEMetric: The Last Stanza of Humanity
  8. ICLR 2026PolyGraph Discrepancy: a classifier-based metric for graph generation
  9. ICLR 2026Positional Encoding Field
  10. ICLR 2026Precise and Interpretable Editing of Code Knowledge in Large Language Models
  11. ICLR 2026PreciseCache: Precise Feature Caching for Efficient and High-fidelity Video Generation
  12. ICLR 2026Programming with Pixels: Can Computer-Use Agents do Software Engineering?
  13. ICLR 2026Prompt-MII: Meta-Learning Instruction Induction for LLMs
  14. ICLR 2026ProReGen: Progressive Residual Generation under Attribute Correlations
  15. ICLR 2026QuRL: Rubrics As Judge For Open-Ended Question Answering
  16. ICLR 2026R1-Code-Interpreter: LLMs Reason with Code via Supervised and Multi-stage Reinforcement Learning
  17. ICLR 2026Rainbow Padding: Mitigating Early Termination in Instruction-Tuned Diffusion LLMs
  18. ICLR 2026RAS: Retrieval-And-Structuring for Knowledge-Intensive LLM Generation
  19. ICLR 2026Reassessing Layer Pruning in LLMs: New Insights and Methods
  20. ICLR 2026RECAST: Expanding the Boundaries of LLMs' Complex Instruction Following with Multi-Constraint Data
  21. ICLR 2026RECODE: A Benchmark for Research Code DEvelopment with Interactive Human Feedback
  22. ICLR 2026RedCodeAgent: Automatic Red-teaming Agent against Diverse Code Agents
  23. ICLR 2026RefineStat: Efficient Exploration for Probabilistic Program Synthesis
  24. ICLR 2026ReFusion: A Diffusion Large Language Model with Parallel Autoregressive Decoding
  25. ICLR 2026Reinforcement Learning from Dynamic Critic Feedback for Free-Form Generations
  26. ICLR 2026Rescue: Retrieval Augmented Secure Code Generation
  27. ICLR 2026Rethinking Code Similarity for Automated Algorithm Design with LLMs
  28. ICLR 2026Rethinking Uncertainty Estimation in LLMs: A Principled Single-Sequence Measure
  29. ICLR 2026Retrospective Sparse Attention for Efficient Long-Context Generation
  30. ICLR 2026ReVeal: Self-Evolving Code Agents via Reliable Self-Verification
  31. ICLR 2026Rewriting Pre-Training Data Boosts LLM Performance in Math and Code
  32. ICLR 2026ROSETTA: Constructing Code-Based Reward from Unconstrained Language Preference
  33. ICLR 2026RPG: A Repository Planning Graph for Unified and Scalable Codebase Generation
  34. ICLR 2026SANA-Video: Efficient Video Generation with Block Linear Diffusion Transformer
  35. ICLR 2026SASFT: Sparse Autoencoder-guided Supervised Finetuning to Mitigate Unexpected Code-Switching in LLMs
  36. ICLR 2026Sat3DGen: Comprehensive Street-Level 3D Scene Generation from Single Satellite Image
  37. ICLR 2026SatDreamer360: Multiview-Consistent Generation of Ground-Level Scenes from Satellite Imagery
  38. ICLR 2026SAVE: A Generalizable Framework for Multi-Condition Single-Cell Generation with Gene Block Attention
  39. ICLR 2026Scalable Energy-Based Models via Adversarial Training: Unifying Discrimination and Generation
  40. ICLR 2026Scalable Training for Vector-Quantized Networks with 100% Codebook Utilization
  41. ICLR 2026Scaling Behavior of Discrete Diffusion Language Models
  42. ICLR 2026Sci2Pol: Evaluating and Fine-tuning LLMs on Scientific-to-Policy Brief Generation
  43. ICLR 2026SciTS: Scientific Time Series Understanding and Generation with LLMs
  44. ICLR 2026SelfReflect: Can LLMs Communicate Their Internal Answer Distribution?
  45. ICLR 2026ShapeGen4D: Towards High Quality 4D Shape Generation from Videos
  46. ICLR 2026ShieldedCode: Learning Robust Representations for Virtual Machine Protected Code
  47. ICLR 2026ShinkaEvolve: Towards Open-Ended and Sample-Efficient Program Evolution
  48. ICLR 2026Shoot First, Ask Questions Later? Building Rational Agents that Explore and Act Like People
  49. ICLR 2026SimpleTIR: End-to-End Reinforcement Learning for Multi-Turn Tool-Integrated Reasoning
  50. ICLR 2026SketchEvo: Leveraging Drawing Dynamics for Enhanced Image Synthesis

在 Claude、Cursor 里直接问这个趋势

接上 Paperlist MCP 后，写综述、挑方向或准备投稿时，不用离开对话框就能拿到这页的数字：每年各会议的论文数、增长倍数、涨得最快的子方向和代表论文，AI 会按你的问题整理。

可以直接复制这句话问 AI

用 paperlist 查代码生成在 NeurIPS 2023–2026 的论文趋势，哪一年涨得最快？

AI 会调用的工具：get_topic_stats {"topic": "code-generation"}

两分钟接入 Paperlist MCP（Claude Code、Cursor、Codex、ChatGPT）

常见问题

2023–2026 年四大会议一共有多少篇代码生成论文？
共 2,161 篇，按年份分别是 2023 年 210 篇、2024 年 446 篇、2025 年 645 篇、2026 年 860 篇。
NeurIPS 2026 有多少篇代码生成论文？
NeurIPS 2026 有 236 篇代码生成论文，上一届是 230 篇。
ICLR 2026 有多少篇代码生成论文？
ICLR 2026 有 241 篇代码生成论文，上一届是 164 篇。
ICML 2026 有多少篇代码生成论文？
ICML 2026 有 174 篇代码生成论文，上一届是 79 篇。
CVPR 2026 有多少篇代码生成论文？
CVPR 2026 有 209 篇代码生成论文，上一届是 172 篇。
哪个会议的代码生成论文最多？
NeurIPS，2023–2026 年共 630 篇。

「自然语言」下的其他主题