大模型基准论文趋势 2023–2026

主要结论

• 大模型基准论文从 2023 年的 2 篇增长到 2026 年的 307 篇，三年间扩大到约 154 倍。
• NeurIPS 是发大模型基准论文最多的会议，2023–2026 年共 256 篇，占四个会议总量的 47%。
• 交叉方向里涨得最快的是AI 智能体：同时属于两个主题的论文从 2023 年的 1 篇增至 2026 年的 116 篇。
• 在「评测与基准」的 5 个主题里，它的 2026 年论文数排第 1；同组增长最快的模型即裁判是 27 倍，大模型基准是 154 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 Free Lunch in LLM Compression: Revisiting Retraining after Pruning
• ICML 2026Controlled LLM Training on Spectral Sphere
• ICLR 2026Hey, That's My Model! Introducing Chain & Hash, An LLM Fingerprinting Technique
• CVPR 20263DrawAgent Teaching LLM to Draw in 3D with Early Contrastive
• NeurIPS 2025ACCO: Accumulate While You Communicate for Communication-Overlapped Sharded LLM Training
• ICML 2025Critical Tokens Matter: Token-Level Contrastive Estimation Enhances LLM’s Reasoning Capability
• ICLR 2025Catastrophic Failure of LLM Unlearning via Quantization
• CVPR 2025Period-LLM: Extending the Periodic Capability of Multimodal Large Language Model
• NeurIPS 2024Large Language Models' Expert-level Global History Knowledge Benchmark (HiST-LLM)
• ICML 2024Characterizing ResNet's Universal Approximation Capability
• ICLR 2024Faithful Explanations of Black-box NLP Models Using LLM-generated Counterfactuals
• CVPR 2024Koala: Key Frame-Conditioned Long Video-LLM
• NeurIPS 2023Quantizable Transformers: Removing Outliers by Helping Attention Heads Do Nothing
• NeurIPS 2026Accelerating LLM Pre-Training through Flat-Direction Dynamics Enhancement
• ICML 2026LLM Priors for ERM over Programs
• ICLR 2026How Learning Rate Decay Wastes Your Best Data in Curriculum-Based LLM Pretraining
• CVPR 20263D-VCD Hallucination Mitigation in 3D-LLM Embodied Agents through Visual Contrastive
• NeurIPS 2025BrainEC-LLM: Brain Effective Connectivity Estimation by Multiscale Mixing LLM
• ICML 2025Position: Medical Large Language Model Benchmarks Should Prioritize Construct Validity
• ICLR 2025Cheating Automatic LLM Benchmarks: Null Models Achieve High Win Rates

全部大模型基准论文

共 544 篇，第 3 / 11 页

101. ICLR 2026Translation Heads: Unveiling Attention's Role in LLM Multilingual Translation
102. ICLR 2026Trinity: An Evolved LLM Coordinator
103. ICLR 2026TSM-Bench: Detecting LLM-Generated Text in Real-World Wikipedia Editing Practices
104. ICLR 2026Universal Model Routing for Efficient LLM Inference
105. ICLR 2026Watch your steps: Dormant Adversarial Behaviors that Activate upon LLM Finetuning
106. ICLR 2026When Greedy Wins: Emergent Exploitation Bias in Meta-Bandit LLM Training
107. ICLR 2026When LLMs get significantly worse: A statistical approach to detect model degradations
108. ICLR 2026WideSearch: Benchmarking Agentic Broad Info-Seeking
109. ICLR 2026WSM: Decay-Free Learning Rate Schedule via Checkpoint Merging for LLM Pre-training
110. ICML 2026$V_0$: A Generalist Value Model for Any Policy at State Zero
111. ICML 2026AdaGC: Enhancing LLM Pretraining Stability via Adaptive Gradient Clipping
112. ICML 2026Adaptive Testing for LLM Evaluation: A Psychometric Alternative to Static Benchmarks
113. ICML 2026Advancing SVD-based LLM Compression via Layer-Wise Error Model Search
114. ICML 2026Beyond Heuristic Tuning: Power-Calibrated LLM Watermarking
115. ICML 2026Beyond Majority Voting: LLM Aggregation by Leveraging Higher-Order Information
116. ICML 2026CapBencher: Give Your LLM Benchmark a Built-in Alarm for Test-Set Overfitting
117. ICML 2026CCLRec: Consensus-driven Contrastive Learning for LLM-enhanced Graph Recommendation
118. ICML 2026Controlled LLM Training on Spectral Sphere
119. ICML 2026CorrSteer: Generation-Time LLM Steering via Correlated Sparse Autoencoder Features
120. ICML 2026CrispEdit: Low-Curvature Projections for Scalable Non-Destructive LLM Editing
121. ICML 2026CriticalKV: Optimizing KV Cache Eviction from an Output Perturbation Perspective
122. ICML 2026Cutting LLM Evaluation Costs with SySRs: A Bandit Algorithm That Provably Exploits Model Similarity
123. ICML 2026Dissecting Quantization Error: A Concentration-Alignment Perspective
124. ICML 2026Distribution-Calibrated Inference Time Compute for Thinking LLM-as-a-Judge
125. ICML 2026DynaSchedBench: Calibrated Dynamic Scheduling Benchmarks and Observability Paradox in LLM-based Scheduling Agents
126. ICML 2026ECHO: Elastic Speculative Decoding with Sparse Gating for High-Concurrency Scenarios
127. ICML 2026EffGen: Enabling Small Language Models as Capable Autonomous Agents
128. ICML 2026Enhancing LLM Training via Spectral Clipping
129. ICML 2026EvoClaw: Evaluating AI Agents on Continuous Software Evolution
130. ICML 2026From Backward Spreading to Forward Replay: Revisiting Target Construction in LLM Parameter Editing
131. ICML 2026GEM: Geometric Entropy Mixing for Optimal LLM Data Curation
132. ICML 2026HexGen-3: A Fully Disaggregated LLM Serving Framework with Fine-Grained Heterogeneous Resource Autoscaling
133. ICML 2026Improving Visual Token Reduction via Rectifying Distortions for Efficient Multimodal LLM Inference
134. ICML 2026Incremental BPE Tokenization
135. ICML 2026LATMiX: Learnable Affine Transformations for Microscaling Quantization of LLMs
136. ICML 2026Learn-to-learn on Arbitrary Textual Conditioning: A Hypernetwork-Driven Meta-gated LLM
137. ICML 2026LLM Priors for ERM over Programs
138. ICML 2026LLM Watermark Evasion via Bias Inversion
139. ICML 2026LLM-based Embeddings: Attention Values Encode Sentence Semantics Better Than Hidden States
140. ICML 2026Minibatch selection for Language Models via Partition Matroid Constrained Gradient Matching
141. ICML 2026Mirror Descent Under Generalized Smoothness
142. ICML 2026OServe: Accelerating LLM Serving via Spatial-Temporal Workload Orchestration
143. ICML 2026OSNIP: Balancing the Privacy-Utility-Efficiency Trilemma in LLM Inference via Obfuscated Semantic Null Space
144. ICML 2026Outrunning LLM Cutoffs: A Live Kernel Crash Resolution Benchmark for All
145. ICML 2026PatternKV: Flattening KV Representation Expands Quantization Headroom
146. ICML 2026PostTrainBench: Can LLM Agents Automate LLM Post-Training?
147. ICML 2026Quantifying Biases in LLM-as-a-Judge Evaluations
148. ICML 2026Quartet II: Accurate LLM Pre-Training in NVFP4 by Improved Unbiased Gradient Estimation
149. ICML 2026QUATRO: Query-Adaptive Trust Region Policy Optimization for LLM Fine-tuning
150. ICML 2026ReSpinQuant: Efficient Layer-Wise LLM Quantization via Subspace Residual Rotation Approximation

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。