Large language model (LLM)-based systems are becoming increasingly popular for solving tasks by constructing executable workflows that interleave LLM calls, information retrieval, tool use, code execution, memory updates, and verification. This survey reviews recent methods for designing and optimizing such workflows, which we treat as agentic computation graphs (ACGs). We classify the literature according to when workflow structure is determined. This lens distinguishes static methods, which fix a reusable workflow scaffold before deployment, from dynamic methods, which select, generate, or revise the workflow for a particular run before or during execution. We further categorize existing work along three dimensions: when structure is determined, which aspect of the workflow is optimized, and which evaluation signals govern optimization (e.g., task metrics, verifier signals, human preferences, or trace-derived feedback). We also distinguish reusable workflow templates, run-specific realized graphs, and execution traces, separating design-time choices from runtime behavior. Finally, we outline a structure-aware evaluation perspective that complements downstream task metrics with graph-level properties, execution cost, robustness, and structural variation across inputs. This work aims to establish a principled vocabulary, a unified framework for situating new methods, a systematic comparison of existing work, and a minimum reporting protocol for LLM agent workflow optimization.

Knowledge graph (KG) embedding is a fundamental task in natural language processing, and various methods have been proposed to explore semantic patterns in distinctive ways. In this paper, we propose to learn an ensemble by leveraging existing methods in a relation-aware manner. However, exploring these semantics using relation-aware ensemble leads to a much larger search space than general ensemble methods. To address this issue, we propose a divide-search-combine algorithm RelEns-DSC that searches the relation-wise ensemble weights independently. This algorithm has the same computation cost as general ensemble methods but with much better performance. Experimental results on benchmark datasets demonstrate the effectiveness of the proposed method in efficiently searching relation-aware ensemble weights and achieving state-of-the-art embedding performance. The code is public at https://github.com/LARS-research/RelEns.