ACL 2025 | 推理不靠堆参数！CRFT打破CoT瓶颈，0.016%参数撬动18.2%性能
"LLM推理" 相关文章
DeltaLLM: A Training-Free Framework Exploiting Temporal Sparsity for Efficient Edge LLM Inference
Revisiting LLM Reasoning via Information Bottleneck
Improving LLMs' Generalized Reasoning Abilities by Graph Problems
Conetext learning 2 KV-cache缓存与命中率
CoLD: Counterfactually-Guided Length Debiasing for Process Reward Models
Accelerating SGLang with Multiple Token Prediction
Agentic Reasoning: A Streamlined Framework for Enhancing LLM Reasoning with Agentic Tools
GHPO: Adaptive Guidance for Stable and Efficient LLM Reinforcement Learning
KisMATH: Do LLMs Have Knowledge of Implicit Structures in Mathematical Reasoning?
On Evaluating Performance of LLM Inference Serving Systems
GraphRunner: A Multi-Stage Framework for Efficient and Accurate Graph-Based Retrieval
Fractional Reasoning in LLMs: A New Way to Control Inference Depth
Krul: Efficient State Restoration for Multi-turn Conversations with Dynamic Cross-layer KV Sharing
Serving LLMs in HPC Clusters: A Comparative Study of Qualcomm Cloud AI 100 Ultra and High-Performance GPUs
Cognitive Load-Aware Inference: A Neuro-Symbolic Framework for Optimizing the Token Economy of Large Language Models
英伟达笑到最后！训练2000步，1.5B逆袭7B巨兽，Scaling真来了
DeepSeek Researchers Open-Sourced a Personal Project named ‘nano-vLLM’: A Lightweight vLLM Implementation Built from Scratch
概率统计机制下，LLM 推理真的「理解世界了」吗？
首次解释LLM如何推理反思！西北大学谷歌新框架：引入贝叶斯自适应强化学习，数学推理全面提升
