Author name: Nikolay Blagoev, O\u{g}uzhan Ersoy, Lydia Yiyu Chen

Hail to the Thief: Exploring Attacks and Defenses in Decentralised GRPO

Nikolay Blagoev, O\u{g}uzhan Ersoy, Lydia Yiyu Chen / April 15, 2026

arXiv:2511.09780v2 Announce Type: replace
Abstract: Group Relative Policy Optimization (GRPO) has demonstrated wide adoption in the post-training of Large Language Models (LLMs). In GRPO, prompts are answered by the model and preferred behaviour is le…

cs.DC, cs.LG

All is Not Lost: LLM Recovery without Checkpoints

Nikolay Blagoev, O\u{g}uzhan Ersoy, Lydia Yiyu Chen / April 7, 2026

arXiv:2506.15461v2 Announce Type: replace-cross
Abstract: Training LLMs on decentralized nodes or on-spot instances, lowers the training cost and enables model democratization. The inevitable challenge here is the transient churns of nodes due to fail…