T${}^{\star }$: Progressive Block Scaling for Masked Diffusion Language Models Through Trajectory Aware Reinforcement Learning

Abstract

We present T${}^{\star }$, a simple TraceRL-based training curriculum for progressive block-size scaling in masked diffusion language models (MDMs). Starting from an AR-initialized small-block MDM, T${}^{\star }$ transitions smoothly to larger blocks, enabling higher-parallelism decoding with minimal performance degradation on math reasoning benchmarks. Moreover, further analysis suggests that T${}^{\star }$ may actually converge to an alternative decoding schedule that achieves comparable performance.