This is a heavily interactive web application, and JavaScript is required. Simple HTML interfaces are possible, but that is not what this is.
Post
askfred.be
askfred.be
did:plc:zugrq46reu5aoeyd2wlb7bv7
The most interesting bit in this paper on offline reinforcement learning for scheduling is how it was trained. It learned effective policies by observing a dataset of completely *random* solutions, not expert demonstrations or suboptimal-but-decent ones. The model found structure in the noise.
https://arxiv.org/abs/2606.11518
2026-06-11T07:01:57.364Z