mlsanityData to 14 Oct 2026

Anticipating the Consequences of Curriculum Decisions with Large Language Models

arXiv 2610.04604 ↗First posted 3 Oct 2026cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Automatic curriculum learning can improve the effectiveness of reinforcement learning by selecting the training experiences presented to the agent over time. Predicting the consequences of such decisions can, however, be difficult. We analyze automatic curriculum learning as a sequential decision-making problem, highlighting a gap between the quantities that determine the value of curriculum decisions and the information captured by local learning signals commonly used to guide them. We then investigate whether Large Language Models (LLMs) can exploit richer information about the learning problem to better anticipate the consequences of curriculum decisions. We introduce a method that combines online learning-progress estimates with LLM-informed estimates of (i) the potential downstream benefits of learning on each task and (ii) whether direct training on a task is currently likely to produce progress. We evaluate the approach on a custom benchmark of 256 textual goals in Craftax under different curriculum objectives. We observe the strongest gains when optimizing for individual target tasks. When optimizing across the full task set, the benefits vary across learners with different mechanisms for cross-task transfer, ranging from modest improvements in learning speed to larger gains that persist through the end of training.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 29 of its 38 references that are in the map. First posted on arXiv 3 Oct 2026.

Builds on

29 of 38 references.

References · 29

  1. 2017Proximal Policy Optimization Algorithmscited by 15,215
  2. 2019Transformer-XL: Attentive Language Models Beyond a Fixed-Length Contextcited by 1,850
  3. 2025Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Modelscited by 1,261
  4. 2017Curiosity-driven Exploration by Self-supervised Predictioncited by 1,194
  5. 2019Solving Rubik's Cube with a Robot Handcited by 734
  6. 2019Stabilizing Transformers for Reinforcement Learningcited by 270
  7. 2019Self-Supervised Exploration via Disagreementcited by 265
  8. 2017Automatic Goal Generation for Reinforcement Learning Agentscited by 259
  9. 2017Teacher-Student Curriculum Learningcited by 210
  10. 2019Skew-Fit: State-Covering Self-Supervised Reinforcement Learningcited by 174
  11. 2020Automatic Curriculum Learning For Deep RL: A Short Surveycited by 150
  12. 2017Intrinsic Motivation and Automatic Curricula via Asymmetric Self-Playcited by 141
17 more
  1. 2019Teacher algorithms for curriculum learning of Deep RL in continuously parameterized environmentscited by 119
  2. 2020Maximum Entropy Gain Exploration for Long Horizon Multi-goal Reinforcement Learningcited by 102
  3. 2024Craftax: A Lightning-Fast Benchmark for Open-Ended Reinforcement Learningcited by 88
  4. 2017Intrinsically Motivated Goal Exploration Processes with Automatic Curriculum Learningcited by 83
  5. 2018CURIOUS: Intrinsically Motivated Modular Multi-Goal Reinforcement Learningcited by 79
  6. 2023OMNI: Open-endedness via Models of human Notions of Interestingnesscited by 70
  7. 2013Active Learning of Inverse Models with Intrinsically Motivated Goal Exploration in Robotscited by 58
  8. 2018Learning Curriculum Policies for Reinforcement Learningcited by 35
  9. 2021Multi-task curriculum learning in a complex, visual, hard-exploration domain: Minecraftcited by 35
  10. 2025SIMA 2: A Generalist Embodied Agent for Virtual Worldscited by 31
  11. 2024CurricuLLM: Automatic Task Curricula Design for Learning Complex Robot Skills using Large Language Modelscited by 22
  12. 2019An Optimization Framework for Task Sequencing in Curriculum Learningcited by 8
  13. 2022Curriculum-based Asymmetric Multi-task Reinforcement Learningcited by 8
  14. 2024Proximal Curriculum with Task Correlations for Deep Reinforcement Learningcited by 7
  15. 2019Curriculum Learning for Cumulative Return Maximizationcited by 4
  16. 2017Faster Reinforcement Learning Using Active Simulatorscited by 2
  17. 2026LEACL: LLM-Enhanced Automatic Curriculum Learning for Reinforcement Learning in Long-Horizon Manipulation Taskscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Self-Evolving Curriculum for LLM Reasoning
  2. 2024On the Modeling Capabilities of Large Language Models for Sequential Decision Making
  3. 2026Actor-Curator: Co-adaptive Curriculum Learning via Policy-Improvement Bandits for RL Post-Training
  4. 2025CurES: From Gradient Analysis to Efficient Curriculum Learning for Reasoning LLMs
  5. 2025Feedback-Induced Performance Decline in LLM-Based Decision-Making
  6. 2025DUMP: Automated Distribution-Level Curriculum Learning for RL-based LLM Post-training
  7. 2025Automating Curriculum Learning for Reinforcement Learning using a Skill-Based Bayesian Network
  8. 2026Internalizing Curriculum Judgment for LLM Reinforcement Fine-Tuning