FlowPIE: Test-Time Scientific Idea Evolution with Flow-Guided Literature Exploration
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866918419677315072 |
|---|---|
| author | Wang, Qiyao Wang, Hongbo Chen, Longze Yang, Zhihao Chen, Guhong Alinejad-Rokny, Hamid Li, Hui Lin, Yuan Yang, Min |
| author_facet | Wang, Qiyao Wang, Hongbo Chen, Longze Yang, Zhihao Chen, Guhong Alinejad-Rokny, Hamid Li, Hui Lin, Yuan Yang, Min |
| contents | Scientific idea generation (SIG) is critical to AI-driven autonomous research, yet existing approaches are often constrained by a static retrieval-then-generation paradigm, leading to homogeneous and insufficiently divergent ideas. In this work, we propose FlowPIE, a tightly coupled retrieval-generation framework that treats literature exploration and idea generation as a co-evolving process. FlowPIE expands literature trajectories via a flow-guided Monte Carlo Tree Search (MCTS) inspired by GFlowNets, using the quality of current ideas assessed by an LLM-based generative reward model (GRM) as a supervised signal to guide adaptive retrieval and construct a diverse, high-quality initial population. Based on this population, FlowPIE models idea generation as a test-time idea evolution process, applying selection, crossover, and mutation with the isolation island paradigm and GRM-based fitness computation to incorporate cross-domain knowledge. It effectively mitigates the information cocoons arising from over-reliance on parametric knowledge and static literature. Extensive evaluations demonstrate that FlowPIE consistently produces ideas with higher novelty, feasibility and diversity compared to strong LLM-based and agent-based frameworks, while enabling reward scaling during test time. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_29557 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | FlowPIE: Test-Time Scientific Idea Evolution with Flow-Guided Literature Exploration Wang, Qiyao Wang, Hongbo Chen, Longze Yang, Zhihao Chen, Guhong Alinejad-Rokny, Hamid Li, Hui Lin, Yuan Yang, Min Artificial Intelligence Computation and Language Scientific idea generation (SIG) is critical to AI-driven autonomous research, yet existing approaches are often constrained by a static retrieval-then-generation paradigm, leading to homogeneous and insufficiently divergent ideas. In this work, we propose FlowPIE, a tightly coupled retrieval-generation framework that treats literature exploration and idea generation as a co-evolving process. FlowPIE expands literature trajectories via a flow-guided Monte Carlo Tree Search (MCTS) inspired by GFlowNets, using the quality of current ideas assessed by an LLM-based generative reward model (GRM) as a supervised signal to guide adaptive retrieval and construct a diverse, high-quality initial population. Based on this population, FlowPIE models idea generation as a test-time idea evolution process, applying selection, crossover, and mutation with the isolation island paradigm and GRM-based fitness computation to incorporate cross-domain knowledge. It effectively mitigates the information cocoons arising from over-reliance on parametric knowledge and static literature. Extensive evaluations demonstrate that FlowPIE consistently produces ideas with higher novelty, feasibility and diversity compared to strong LLM-based and agent-based frameworks, while enabling reward scaling during test time. |
| title | FlowPIE: Test-Time Scientific Idea Evolution with Flow-Guided Literature Exploration |
| topic | Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2603.29557 |