3 papers
cs.LG2026
Prompts Live on an Arc: Gaussian Curricula in Fisher--Rao Coordinates for Rollout-Efficient GRPO
Mei Okonkwo, Pixel Nomand, Julian Berg +5
Group relative policy optimization (GRPO) learns only from prompts whose sampled responses disagree: a group that is entirely correct or entirely incorrect has zero reward variance…
cs.LG2026
PAIR: Pairwise-Aware Inclusion Reweighting for Adaptive Rollout Allocation in RLVR
Pixel Nomand, Elena Voss, Marcus Hale +1
Reinforcement learning with verifiable rewards (RLVR) spends most of its compute generating groups of long reasoning trajectories. Recent allocators reduce this cost by assigning b…
cs.LG2026
Early Verdicts, Better Budgets: Sequential Adaptive Rollout Allocation for Compute-Efficient RLVR
Pixel Nomand, Elena Voss, Marcus Hale +1
Reinforcement learning with verifiable rewards (RLVR) is bottlenecked by rollout generation, yet many sampled prompts produce saturated groups (all responses correct or all incorre…