2 papers
cs.AI2026
Depth as a Candidate Source: Portfolio Allocation for Recurrent Test-Time Scaling in Large Language Models
Chaoyuan Hao, Wentao Yue, Yanyan Wu +6
Recurrent latent reasoning in large lanugage models (LLMs) scales test-time computation by increasing recurrent depth, naturally framing inference as a problem of deciding when to…
cs.AR2026
A Per-Step Load-Balancing Engine for Expert-Parallel MoE Serving
Jingyuan Li, Maolin Wang, Hayden So
In expert-parallel Mixture-of-Experts serving, a layer finishes only when the most loaded GPU of its group does. Token routing shifts at every decode step and concentrates on a few…