A Study of the Efficacy of Generative Flow Networks for Robotics and Machine Fault-Adaptation Journal: Engineering Application of Artificial Intelligence
Zahin Sufiyan, Shadan Golestan, Shotaro Miwa, Yoshihiro Mitsuka, Osmar Zaiane · 2025
Read A Study of the Efficacy of Generative Flow Networks for Robotics and Machine Fault-Adaptation Journal: Engineering Application of Artificial Intelligence by Zahin Sufiyan, Shadan Golestan, Shotaro Miwa, Yoshihiro Mitsuka, Osmar Zaiane (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-03405.txt.
Description
Zahin Sufiyan and co-authors test, in a 2025 robotics paper, whether a sampling-based learning method can teach machines to adapt fast when hardware faults strike. Faults are a top real-world surprise that strands robots trained for normal conditions. The team used a simulated reaching arm and broke it in four distinct ways.
Their method, a continuous form of generative flow networks, learns to sample good actions in proportion to reward. It grew adaptive behavior under the fault conditions. Head-to-head runs against standard reinforcement learning showed where each approach wins on adaptation speed and data thrift. The authors frame flow networks as a route to fault-tolerant machines.
Catalog record
| CREATOR | Zahin Sufiyan, Shadan Golestan, Shotaro Miwa, Yoshihiro Mitsuka, Osmar Zaiane |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 90 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-03405