GLOW: Graph-Language Co-Reasoning for Agentic Workflow Performance Prediction

TL;DR

GLOW is a novel framework combining graph neural networks and large language models to improve agentic workflow performance prediction by capturing complex topological and semantic features, leading to superior accuracy.

Contribution

It introduces a graph-oriented LLM and a contrastive alignment strategy to enhance AW performance prediction, addressing limitations of existing methods.

Findings

GLOW outperforms state-of-the-art baselines in accuracy.

GLOW achieves better ranking utility.

The framework effectively captures topological and semantic features.

Abstract

Agentic Workflows (AWs) have emerged as a promising paradigm for solving complex tasks. However, the scalability of automating their generation is severely constrained by the high cost and latency of execution-based evaluation. Existing AW performance prediction methods act as surrogates but fail to simultaneously capture the intricate topological dependencies and the deep semantic logic embedded in AWs. To address this limitation, we propose GLOW, a unified framework for AW performance prediction that combines the graph-structure modeling capabilities of GNNs with the reasoning power of LLMs. Specifically, we introduce a graph-oriented LLM, instruction-tuned on graph tasks, to extract topologically aware semantic features, which are fused with GNN-encoded structural representations. A contrastive alignment strategy further refines the latent space to distinguish high-quality AWs. Extensive experiments on FLORA-Bench show that GLOW outperforms state-of-the-art baselines in prediction accuracy and ranking utility.