Toward Reliable Design of LLM-Enabled Agentic Workflows: Optimizing Latency-Reliability-Cost Tradeoffs 文章

ArXiv CS.AI2026-05-26NEWSen作者: Ya-Ting Yang, Quanyan Zhu
大语言模型
查看原文 →
关系图谱

Toward Reliable Design of LLM-Enabled Agentic Workflows: Optimizing Latency-Reliability-Cost Tradeoffs · 相关技术

相关技术

LLMORMODElanguage modeldivide-and-conquer partitioningTokenTeraStraight-Through EstimatorSULParts-of-Speech (POS) tagsOWSNarrative Abstraction BenchmarkLMMlarge language modelsLLM agentsHISForENAARGANNAgentICETPU