Dynamic Upcycling MoE (DUME) cleverly reuses dense experts trained on different domains to create a unified MoE multidomain expert model. DUME retains the knowledge of the original dense experts without any additional training, offering a cost-effective, scalable solution with no compromises.

In this post we'll walk through the inner-workings of prediction markets and the systems they exist inside of.

We present a decentralised two-sided market design that treats compute as a time‑bound asset, enabled by reproducibility, verification, and checkpointing, yielding dynamic pricing and simple matching without combinatorial auctions.

Our paper, “Hail to the Thief: Exploring Attacks and Defenses in Decentralized GRPO”, is the first systematic study that explores both the attack vectors and defense strategies in decentralised reinforcement learning for Large Language Models (LLMs).

This is an academic paper describing SAPO, a meta-algorithm that wraps around your preferred policy gradient algorithm.