Large Hadron Collider
PulseAugur coverage of Large Hadron Collider — every cluster mentioning Large Hadron Collider across labs, papers, and developer communities, ranked by signal.
4 天有情绪数据
-
Hyper-Graph Neural Networks enhance LHC particle collision analysis
Researchers have developed a Hyper-Graph Neural Network (H-GNN) to improve the detection of $tar{t}tar{t}$ production at the Large Hadron Collider. This advanced neural network architecture represents events as hyperg…
-
New neural inference method targets Higgs self-coupling at LHC
Researchers have developed a novel neural simulation-based inference (NSBI) approach to determine the Higgs trilinear self-coupling. This method combines the efficiency of matrix-element-enhanced techniques with the pra…
-
LLM agents struggle with scientific reasoning; Cerebras IPO challenges Nvidia
A new benchmark, Collider-Bench, has been developed to evaluate the ability of large language model agents to reproduce scientific analyses from research papers, specifically focusing on Large Hadron Collider (LHC) data…
-
新型SNAC-Pack为FPGA自动化神经架构协同设计
研究人员开发了SNAC-Pack，这是一个开源框架，旨在自动化神经架构的协同设计及其在FPGA上的部署。该软件包采用多目标全局搜索策略，并结合硬件代理模型来估算资源使用和延迟，从而在搜索过程中避免昂贵的综合。SNAC-Pack已被证明在发现用于大型强子对撞机中的喷流分类和超导量子比特读出等任务的紧凑型架构方面非常有效，显著缩短了设计探索时间，同时保持或提高了性能。
-
Transfer learning boosts AI model efficiency in high-energy physics
Researchers have explored transfer learning techniques to improve machine learning model performance in high-energy physics. By pre-training models on computationally cheaper, fast-simulated data and then adapting them …
-
AI研究人员比较粒子物理学中喷注鉴别方法的可解释性
研究人员开发并比较了三种可解释人工智能（XAI）方法——GNNExplainer、GNNShap 和 GradCAM——以理解在大型强子对撞机喷注鉴别中使用的图神经网络的预测。该研究将这些XAI技术应用于Lund平面表示，该表示将部分子分裂映射到图节点。通过引入一个物理信息评估框架，该研究量化了不同能量区域下解释质量的变化，并评估了AI分配的重要性与已建立的喷注子结构可观测值之间的相关性。
-
jBOT uses self-distillation to cluster jet representations for LHC data
Researchers have developed jBOT, a novel self-supervised learning method for analyzing particle physics data from the CERN Large Hadron Collider. This technique utilizes self-distillation, combining local and global dis…
-
New HGQ-LUT and da4ml methods speed up DNN training and FPGA deployment
Researchers have developed HGQ-LUT, a new method for training lookup-table (LUT) based neural networks that significantly speeds up the training process, making it over 100 times faster on modern GPUs. This approach int…