Blogs

2026

• IsoCompute Playbook: Optimally Scaling Sampling Compute for RL Training of LLMs
  Z. Cheng^✱, Y. Xie^✱, Y. Qu^✱, ^✱, S. Hao, V. Pimpalkhute, T. Liang, F. Yao, Z. Liu, E. Xing, V. Smith, R. Salakhutdinov, Z. Hu, T. Killian, A. Kumar
  [CMU MLD Blog] (Jan 2026)

2025

• How to Explore to Scale RL Training of LLMs on Hard Problems?
  Y. Qu^✱, ^✱, V. Smith, R. Salakhutdinov, A. Kumar
  [Blog] (Dec 2025)

• Sharpening or Discovery, RL or Meta RL?: How RL Improves LLM Reasoning
  , A. Kumar
  [Blog] (June 2025)

• Optimizing LLM test-time compute involves solving a meta RL problem
  , Y. Qu, M. Yang, L. Zhang, V. Smith, A. Kumar
  [CMU MLD Blog] (Jan 2025)