Showing cs.LGShow all
3 papers · 1 filter
cs.LG2026
Mechanism Shift During Post-training from Autoregressive to Masked Diffusion Language Models
Injin Kong, Hyoungjoon Lee, Yohan Jo
Post-training pretrained autoregressive models (ARMs) into masked diffusion models (MDMs) has emerged as a cost-effective way to overcome the limitations of sequential generation.…
cs.LG2026
Your Language Model is Its Own Critic: Reinforcement Learning with Value Estimation from Actor's Internal States
Yunho Choi, Jongwon Lim, Woojin Ahn +3
Reinforcement learning with verifiable rewards (RLVR) for Large Reasoning Models hinges on baseline estimation for variance reduction, but existing approaches pay a heavy price: PP…
cs.LG2026
KL for a KL: On-Policy Distillation with Control Variate Baseline
Minjae Oh, Sangjun Song, Gyubin Choi +2
On-Policy Distillation (OPD) has emerged as a dominant post-training paradigm for large language models, especially for reasoning domains. However, OPD remains unstable in practice…