GLOW: Graph-Language Co-Reasoning for Agentic Workflow Performance Prediction
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911324984836096 |
|---|---|
| author | Guan, Wei Cao, Jian Cai, Jinyu Cai, Qiqi Gao, Jianqi Ng, See-Kiong |
| author_facet | Guan, Wei Cao, Jian Cai, Jinyu Cai, Qiqi Gao, Jianqi Ng, See-Kiong |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_15751 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | GLOW: Graph-Language Co-Reasoning for Agentic Workflow Performance Prediction Guan, Wei Cao, Jian Cai, Jinyu Cai, Qiqi Gao, Jianqi Ng, See-Kiong Machine Learning Artificial Intelligence Multiagent Systems 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. |
| title | GLOW: Graph-Language Co-Reasoning for Agentic Workflow Performance Prediction |
| topic | Machine Learning Artificial Intelligence Multiagent Systems |
| url | https://arxiv.org/abs/2512.15751 |