Feature
Empirical Sparse-to-Dense Reward Principle for LM Post-Training
Amy Reed · 13 May 2026
Researchers propose a sparse-to-dense reward principle for language model post-training when labeled data is scarce. It outperforms standard GRPO and on-policy distillation by allocating checked examples efficiently.
Researchers introduced the sparse-to-dense reward principle for language model post-training. This method prioritizes careful allocation of scarce labeled verifiable data over standard GRPO on deployment models. Empirical findings show it outperforms GRPO and on-policy distillation in data-constrained settings.