New LLM RL paper says PPO-Clip hurts exploration and RIPO lifts AIME24 by 60%
burny_tech · x · 2026-07-25
A new paper argues that PPO-Clip in LLM reinforcement learning has a geometric flaw: it measures policy change in Euclidean terms even though policy updates live on a Riemannian manifold.
The authors say this mismatch suppresses rare but useful actions, causing exploration collapse. Their fix, Riemannian Isometric Policy Optimization (RIPO), gives low-probability actions more room and high-probability actions less, improving stability and reportedly boosting AIME24 performance by up to 60% over GRPO across seven benchmarks.
Related event: RIPO Overcomes PPO-Clip's Exploration Collapse in LLM RL(2 posts)→
More from Research
- Statistical physics paper studies optimal MLP learning near interpolation — burny_tech · 2026-07-25
- ICML paper says regularized learning often looks Hebbian, while noise turns it anti-Hebbian — burny_tech · 2026-07-25
- AI Autonomously Generates 9,000 Lines of Math Proofs for Fluid Dynamics — burny_tech · 2026-07-25
- Could rare punctuation marks become one-token tone signals for LLMs? — Fcking_Chuck · 2026-07-25
- RL Consistently Improves Imagination Models: Photon-1 Beats Gemini — ycombinator · 2026-07-25
- A user wants one-photo side and back views without changing pose for 3D modeling — LoudPoem9492 · 2026-07-25