An enterprise consolidated 200+ internal apps onto one self-hosted model by training three separate GRPO experts and merging them with two-stage SLERP

arXiv·medium signalXBlueskyLinkedInCopy link

arXiv 2609.01572 (2026-09-01, 33 HF upvotes) is a rare production write-up rather than a benchmark paper: data-residency rules force self-hosting, and stacking newer models without retiring old ones fragments a finite GPU pool. The team closed quality gaps along instruction following, function calling and internal task distribution, but found that optimizing all three jointly caused cross-domain reward interference, so they trained one GRPO expert per axis and merged via two-stage SLERP. Each reward exposed a distinct failure mode with its own fix (semantic collapse, over-calling, verbosity hacking), and in non-reasoning mode the merged model beats one roughly 7x larger by total parameters.