爱可可 AI 前沿推介(8.2)

文章预览

LG - 机器学习 CV - 计算机视觉 CL - 计算与语言 1、[LG] Large Language Monkeys：Scaling Inference Compute with Repeated Sampling 2、[CL] Dancing in Chains：Reconciling Instruction Following and Faithfulness in Language Models 3、[LG] Deep Learning for Options Trading：An End-To-End Approach 4、[LG] MoMa：Efficient Early-Fusion Pre-training with Mixture of Modality-Aware Experts 5、[CL] Towards Achieving Human Parity on End-to-end Simultaneous Speech Translation via LLM Agent 摘要：用重复采样扩展推理计算、语言模型中的指令遵循与忠实性、基于深度学习的期权交易端到端方法、模态感知专家混合的高效早期融合预训练、基于LLM Agent的端到端同步语音翻译 1、[LG] Large Language Monkeys: Scaling Inference Compute with Repeated Sampling B Brown, J Juravsky, R Ehrlich, R Clark… [Stanford University & University of Oxford & Google DeepMind] 大型语言猴子：用重复采样扩展推理计算 要点: ………………………………