SAGC dynamically adjusts group sizes in synchronous GRPO and DAPO via online constrained optimization to cut stragglers, improve wall-clock speed, and maintain or improve rewards and downstream reasoning performance.
Efficient rlvr training via weighted mutual information data selection.arXiv preprint arXiv:2603.01907
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.LG 3years
2026 3verdicts
UNVERDICTED 3representative citing papers
ATLAS traces RLVR data to 20 atomic sources, most datasets are variants, and DAPO++ curated with SCA improves RLVR performance while Q predicts training effectiveness.
NExt accelerates RLVR training for LLMs by nonlinearly extrapolating low-rank parameter trajectories extracted from LoRA runs.
citing papers explorer
-
Faster Synchronous On-Policy RL via Straggler-Aware Group Sizing
SAGC dynamically adjusts group sizes in synchronous GRPO and DAPO via online constrained optimization to cut stragglers, improve wall-clock speed, and maintain or improve rewards and downstream reasoning performance.
-
RLVR Datasets and Where to Find Them: Tracing Data Lineage for Better Training Data
ATLAS traces RLVR data to 20 atomic sources, most datasets are variants, and DAPO++ curated with SCA improves RLVR performance while Q predicts training effectiveness.
-
Low-rank Optimization Trajectories Modeling for LLM RLVR Acceleration
NExt accelerates RLVR training for LLMs by nonlinearly extrapolating low-rank parameter trajectories extracted from LoRA runs.