GNNs as Predictors of Agentic Workflow Performances
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910875852472320 |
|---|---|
| author | Zhang, Yuanshuo Hou, Yuchen Tang, Bohan Chen, Shuo Zhang, Muhan Dong, Xiaowen Chen, Siheng |
| author_facet | Zhang, Yuanshuo Hou, Yuchen Tang, Bohan Chen, Shuo Zhang, Muhan Dong, Xiaowen Chen, Siheng |
| contents | Agentic workflows invoked by Large Language Models (LLMs) have achieved remarkable success in handling complex tasks. However, optimizing such workflows is costly and inefficient in real-world applications due to extensive invocations of LLMs. To fill this gap, this position paper formulates agentic workflows as computational graphs and advocates Graph Neural Networks (GNNs) as efficient predictors of agentic workflow performances, avoiding repeated LLM invocations for evaluation. To empirically ground this position, we construct FLORA-Bench, a unified platform for benchmarking GNNs for predicting agentic workflow performances. With extensive experiments, we arrive at the following conclusion: GNNs are simple yet effective predictors. This conclusion supports new applications of GNNs and a novel direction towards automating agentic workflow optimization. All codes, models, and data are available at https://github.com/youngsoul0731/Flora-Bench. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_11301 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | GNNs as Predictors of Agentic Workflow Performances Zhang, Yuanshuo Hou, Yuchen Tang, Bohan Chen, Shuo Zhang, Muhan Dong, Xiaowen Chen, Siheng Computation and Language Multiagent Systems Agentic workflows invoked by Large Language Models (LLMs) have achieved remarkable success in handling complex tasks. However, optimizing such workflows is costly and inefficient in real-world applications due to extensive invocations of LLMs. To fill this gap, this position paper formulates agentic workflows as computational graphs and advocates Graph Neural Networks (GNNs) as efficient predictors of agentic workflow performances, avoiding repeated LLM invocations for evaluation. To empirically ground this position, we construct FLORA-Bench, a unified platform for benchmarking GNNs for predicting agentic workflow performances. With extensive experiments, we arrive at the following conclusion: GNNs are simple yet effective predictors. This conclusion supports new applications of GNNs and a novel direction towards automating agentic workflow optimization. All codes, models, and data are available at https://github.com/youngsoul0731/Flora-Bench. |
| title | GNNs as Predictors of Agentic Workflow Performances |
| topic | Computation and Language Multiagent Systems |
| url | https://arxiv.org/abs/2503.11301 |