Researchers Trace RLVR's Roots: It's Just RL, and the Idea Isn't New

AI researchers debated the conceptual history of RLVR, with lateinteraction arguing it is essentially just RL and that self-training LLMs to use tools predates 2021, while soldni noted the idea existed but lacked a widely usable recipe.

2026-10-15 ~ 2026-10-15 · 3 related posts

• Tool use is older than you think: LaMDA paper casually mentioned it in Jan 2022 — soldni · 2026-10-15
• RLVR is just RL — RLHF is the weird one, argues lateinteraction — lateinteraction · 2026-10-15
• Researchers debate RLVR's lineage: 'RLHF was the weird one, STaR was already around' — soldni · 2026-10-15