Actor-Curator: Co-adaptive Curriculum Learning via Policy-Improvement Bandits for RL Post-Training

• URL: http://arxiv.org/abs/2602.20532v1
• Date: Tue, 24 Feb 2026 04:19:48 GMT
• Title: Actor-Curator: Co-adaptive Curriculum Learning via Policy-Improvement Bandits for RL Post-Training
• Authors: Zhengyao Gu, Jonathan Light, Raul Astudillo, Ziyu Ye, Langzhou He, Henry Peng Zou, Wei Cheng, Santiago Paternain, Philip S. Yu, Yisong Yue,
• Abstract summary: ACTOR-CURATOR is a scalable and fully automated curriculum learning framework for reinforcement learning post-training of large language models.<n> Empirically, ACTOR-CURATOR consistently outperforms uniform sampling and strong curriculum baselines.
• Score: 63.34044358216334
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Post-training large foundation models with reinforcement learning typically relies on massive and heterogeneous datasets, making effective curriculum learning both critical and challenging. In this work, we propose ACTOR-CURATOR, a scalable and fully automated curriculum learning framework for reinforcement learning post-training of large language models (LLMs). ACTOR-CURATOR learns a neural curator that dynamically selects training problems from large problem banks by directly optimizing for expected policy performance improvement. We formulate problem selection as a non-stationary stochastic bandit problem, derive a principled loss function based on online stochastic mirror descent, and establish regret guarantees under partial feedback. Empirically, ACTOR-CURATOR consistently outperforms uniform sampling and strong curriculum baselines across a wide range of challenging reasoning benchmarks, demonstrating improved training stability and efficiency. Notably, it achieves relative gains of 28.6% on AIME2024 and 30.5% on ARC-1D over the strongest baseline and up to 80% speedup. These results suggest that ACTOR-CURATOR is a powerful and practical approach for scalable LLM post-training.

Related papers

• Buffer Matters: Unleashing the Power of Off-Policy Reinforcement Learning in Large Language Model Reasoning [12.863583402455008]
  Batch Adaptation Policy Optimization (BAPO) is an off-policy RLVR framework to improve the data efficiency in large language models post-training.<n>It dynamically selects training batches by re-evaluating historically difficult samples and reusing high-quality ones.<n>BAPO achieves an average 12.5% improvement over GRPO across mathematics, planning, and visual reasoning tasks.
  arXiv  Detail & Related papers  (2026-02-24T09:35:43Z)
• Contextual Rollout Bandits for Reinforcement Learning with Verifiable Rewards [69.74686029941881]
  Reinforcement Learning with Verifiable Rewards (RLVR) is an effective paradigm for improving the reasoning capabilities of large language models.<n>We propose a unified neural scheduling framework that adaptively selects high-value rollouts throughout training.<n>Experiments on six mathematical reasoning benchmarks demonstrate consistent gains in performance and training efficiency across multiple RLVR optimization methods.
  arXiv  Detail & Related papers  (2026-02-09T10:51:58Z)
• Resource-Efficient Reinforcement for Reasoning Large Language Models via Dynamic One-Shot Policy Refinement [21.073482007189504]
  Large language models (LLMs) have exhibited remarkable performance on complex reasoning tasks.<n> reinforcement learning under verifiable rewards (RLVR) is emerging as a principled framework for aligning model behavior with reasoning chains.<n>Despite its promise, RLVR remains prohibitively resource-intensive, requiring extensive reward signals and incurring substantial rollout costs during training.
  arXiv  Detail & Related papers  (2026-01-31T16:51:50Z)
• Human-in-the-loop Online Rejection Sampling for Robotic Manipulation [55.99788088622936]
  Hi-ORS stabilizes value estimation by filtering out negatively rewarded samples during online fine-tuning.<n>Hi-ORS fine-tunes a pi-base policy to master contact-rich manipulation in just 1.5 hours of real-world training.
  arXiv  Detail & Related papers  (2025-10-30T11:53:08Z)
• Beyond Pass@1: Self-Play with Variational Problem Synthesis Sustains RLVR [110.90317717368264]
  We propose an online Self-play with Variational problem Synthesis (SvS) strategy for RLVR training.<n>This strategy effectively maintains policy entropy during training and substantially improves Pass@k compared with standard RLVR.
  arXiv  Detail & Related papers  (2025-08-19T17:42:45Z)
• Actor-Critic for Continuous Action Chunks: A Reinforcement Learning Framework for Long-Horizon Robotic Manipulation with Sparse Reward [85.84943447589511]
  This paper introduces AC3 (Actor-Critic for Continuous Chunks), a novel RL framework that learns to generate high-dimensional, continuous action sequences.<n>To make this learning process stable and data-efficient, AC3 incorporates targeted stabilization mechanisms for both the actor and the critic.
  arXiv  Detail & Related papers  (2025-08-15T01:27:15Z)
• GHPO: Adaptive Guidance for Stable and Efficient LLM Reinforcement Learning [15.43938821214447]
  Reinforcement Learning with Verifiable Rewards (RLVR) has recently emerged as a powerful paradigm for facilitating the self-improvement of large language models (LLMs)<n>This paper introduces Guided Hybrid Policy Optimization (GHPO), a novel difficulty-aware reinforcement learning framework.<n>GHPO dynamically calibrates task difficulty by employing adaptive prompt refinement to provide targeted guidance.
  arXiv  Detail & Related papers  (2025-07-14T08:10:00Z)
• Online Difficulty Filtering for Reasoning Oriented Reinforcement Learning [8.537540092998311]
  Reasoning-Oriented Reinforcement Learning (RORL) enhances the reasoning ability of Large Language Models (LLMs)<n>We show that curating the batch with the problems that the training model achieves intermediate accuracy on the fly can maximize the effectiveness of RORL training.
  arXiv  Detail & Related papers  (2025-04-04T11:52:05Z)
• Dynamic Learning Rate for Deep Reinforcement Learning: A Bandit Approach [2.743898388459522]
  In deep Reinforcement Learning (RL), the learning rate critically influences both stability and performance, yet its optimal value shifts during training as the environment and policy evolve.<n>Standard decay schedulers assume monotonic convergence and often misalign with these dynamics, leading to premature or delayed adjustments.<n>We introduce LRRL, a meta-learning approach that dynamically selects the learning rate based on policy performance rather than training steps.
  arXiv  Detail & Related papers  (2024-10-16T14:15:28Z)
• Training Language Models to Self-Correct via Reinforcement Learning [98.35197671595343]
  Self-correction has been found to be largely ineffective in modern large language models (LLMs) We develop a multi-turn online reinforcement learning approach, SCoRe, that significantly improves an LLM's self-correction ability using entirely self-generated data. We find that SCoRe achieves state-of-the-art self-correction performance, improving the base models' self-correction by 15.6% and 9.1% respectively on MATH and HumanEval.
  arXiv  Detail & Related papers  (2024-09-19T17:16:21Z)
• Mastering the Unsupervised Reinforcement Learning Benchmark from Pixels [112.63440666617494]
  Reinforcement learning algorithms can succeed but require large amounts of interactions between the agent and the environment. We propose a new method to solve it, using unsupervised model-based RL, for pre-training the agent. We show robust performance on the Real-Word RL benchmark, hinting at resiliency to environment perturbations during adaptation.
  arXiv  Detail & Related papers  (2022-09-24T14:22:29Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.