大模型基准论文趋势 2023–2026

能力评测

统计范围：NeurIPS、ICLR、ICML、CVPR 四个会议，2023–2026 年，共 544 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

大模型基准论文数，按年份和会议
年份NeurIPSICLRICMLCVPR合计
202320002
2024241321361
20259250248174
2026138103606307

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

全部大模型基准论文

年份全部2023202420252026
会议全部NeurIPSICLRICMLCVPR

共 544 篇，第 7 / 11 页

  1. NeurIPS 2026We Need to Improve Benchmarks in AI for Mathematics
  2. NeurIPS 2026What Claims Do LLM Benchmark Scores Support?
  3. NeurIPS 2026What Should Remain After Forgetting? Rethinking LLM Unlearning as Predictive Posterior Correction
  4. NeurIPS 2026When Does Online Imitation Learning Help in LLM Post-Training? The Role of (Non-)Realizability Beyond Horizon
  5. NeurIPS 2026When Expert Disagreement Hurts: Auditing Prestige-Sensitive Revision in LLM Decision Pipelines
  6. NeurIPS 2026When LLM Routers Overpay: Strong-Model Over-Selection under Loose Budgets
  7. NeurIPS 2026XBRIDGE: Entity-Grounded Latent Bridge for Heterogeneous LLM Communication
  8. CVPR 2025LiveCC: Learning Video LLM with Streaming Speech Transcription at Scale
  9. CVPR 2025M-LLM Based Video Frame Selection for Efficient Video Understanding
  10. CVPR 2025Period-LLM: Extending the Periodic Capability of Multimodal Large Language Model
  11. CVPR 2025Separation of Powers: On Segregating Knowledge from Observation in LLM-enabled Knowledge-based Visual Question Answering
  12. CVPR 2025StoryGPT-V: Large Language Models as Consistent Story Visualizers
  13. CVPR 2025Video-3D LLM: Learning Position-Aware Video Representation for 3D Scene Understanding
  14. CVPR 2025VideoRefer Suite: Advancing Spatial-Temporal Object Understanding with Video LLM
  15. CVPR 2025VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos
  16. ICLR 2025AI Sandbagging: Language Models can Strategically Underperform on Evaluations
  17. ICLR 2025An Empirical Analysis of Uncertainty in Large Language Model Evaluations
  18. ICLR 2025Better than Your Teacher: LLM Agents that learn from Privileged AI Feedback
  19. ICLR 2025BingoGuard: LLM Content Moderation Tools with Risk Levels
  20. ICLR 2025Catastrophic Failure of LLM Unlearning via Quantization
  21. ICLR 2025Cheating Automatic LLM Benchmarks: Null Models Achieve High Win Rates
  22. ICLR 2025Combatting Dimensional Collapse in LLM Pre-Training Data via Submodular File Selection
  23. ICLR 2025Controlled LLM Decoding via Discrete Auto-regressive Biasing
  24. ICLR 2025Cut the Crap: An Economical Communication Pipeline for LLM-based Multi-Agent Systems
  25. ICLR 2025Determine-Then-Ensemble: Necessity of Top-k Union for Large Language Model Ensembling
  26. ICLR 2025Diversity Empowers Intelligence: Integrating Expertise of Software Engineering Agents
  27. ICLR 2025Do LLMs have Consistent Values?
  28. ICLR 2025Do LLMs Recognize Your Preferences? Evaluating Personalized Preference Following in LLMs
  29. ICLR 2025DON’T STOP ME NOW: EMBEDDING BASED SCHEDULING FOR LLMS
  30. ICLR 2025ELICIT: LLM Augmentation Via External In-context Capability
  31. ICLR 2025Faster Cascades via Speculative Decoding
  32. ICLR 2025ForecastBench: A Dynamic Benchmark of AI Forecasting Capabilities
  33. ICLR 2025Fourier Head: Helping Large Language Models Learn Complex Probability Distributions
  34. ICLR 2025From Commands to Prompts: LLM-based Semantic File System for AIOS
  35. ICLR 2025GameArena: Evaluating LLM Reasoning through Live Computer Games
  36. ICLR 2025Herald: A Natural Language Annotated Lean 4 Dataset
  37. ICLR 2025HShare: Fast LLM Decoding by Hierarchical Key-Value Sharing
  38. ICLR 2025Improving Large Language Model Planning with Action Sequence Similarity
  39. ICLR 2025Instruct-SkillMix: A Powerful Pipeline for LLM Instruction Tuning
  40. ICLR 2025Instructional Segment Embedding: Improving LLM Safety with Instruction Hierarchy
  41. ICLR 2025Language Models Are Implicitly Continuous
  42. ICLR 2025Large Language Models can Become Strong Self-Detoxifiers
  43. ICLR 2025Law of the Weakest Link: Cross Capabilities of Large Language Models
  44. ICLR 2025Lawma: The Power of Specialization for Legal Annotation
  45. ICLR 2025LiveBench: A Challenging, Contamination-Limited LLM Benchmark
  46. ICLR 2025LLM Unlearning via Loss Adjustment with Only Forget Data
  47. ICLR 2025Logical Consistency of Large Language Models in Fact-Checking
  48. ICLR 2025MM1.5: Methods, Analysis & Insights from Multimodal LLM Fine-tuning
  49. ICLR 2025Omni-MATH: A Universal Olympiad Level Mathematic Benchmark for Large Language Models
  50. ICLR 2025On Calibration of LLM-based Guard Models for Reliable Content Moderation

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 查大模型基准在 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": "llm-benchmarks"}

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大模型基准论文？
共 544 篇，按年份分别是 2023 年 2 篇、2024 年 61 篇、2025 年 174 篇、2026 年 307 篇。
NeurIPS 2026 有多少篇大模型基准论文？
NeurIPS 2026 有 138 篇大模型基准论文，上一届是 92 篇。
ICLR 2026 有多少篇大模型基准论文？
ICLR 2026 有 103 篇大模型基准论文，上一届是 50 篇。
ICML 2026 有多少篇大模型基准论文？
ICML 2026 有 60 篇大模型基准论文，上一届是 24 篇。
CVPR 2026 有多少篇大模型基准论文？
CVPR 2026 有 6 篇大模型基准论文，上一届是 8 篇。
哪个会议的大模型基准论文最多？
NeurIPS，2023–2026 年共 256 篇。

「评测与基准」下的其他主题