AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

On the Emergence of Implicit Curriculum in RLVR Learning Dynamics

Yu Huang, Zixin Wen, Yuejie Chi, Yuting Wei, Aarti Singh, Yingbin Liang, Yuxin Chen

ICML 2026 · Reinforcement learning › RL from feedback · approach: Scaling RL with verifiable rewards · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
train with outcome-only RL on a short horizon (L = 5) (ablation)one-layer attention with frozen pretrained MLP, REINFORCE · synthetic Z96 state tracking, fixed lengthreward and attention concentrationImproved
train with outcome-only RL on longer fixed horizons (L = 15, 45) (ablation)one-layer attention with frozen pretrained MLP, REINFORCE · synthetic Z96 state tracking, fixed lengthreward and attention concentrationHurt
mix horizons with moderate difficulty ratio R = 3 (L in {5,15,45}) (ablation)mixed-difficulty REINFORCE · synthetic Z96 state tracking, mixed lengthreward trajectory across horizonsImproved
mix horizons with large difficulty ratio R = 7 (L in {5,35}) (ablation)mixed-difficulty REINFORCE · synthetic Z96 state tracking, mixed lengthreward trajectory across horizonsHurt
replace outcome-based RL with supervised fine-tuning with step-level supervision (theory)one-layer transformer · long-horizon composition beyond critical horizonlearnabilityImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy