The paper investigates how curriculum difficulty progression affects training in Group Relative Policy Optimization (GRPO). It finds that continually selecting medium‑difficulty samples is not optimal; a staged curriculum that gradually shifts toward harder samples improves performance, reaching a peak accuracy of 0.6591. However, progressing too far to very hard samples degrades accuracy, indicating a non‑monotonic benefit of difficulty progression.
arXiv:2603. 13761v2 Announce Type: replace Abstract: Curriculum learning--ordering training examples in a sequence to aid machine learning--takes inspiration from human learning, but has not gained widespread acceptance.
arXiv:2609.40115v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has become an important approach for improving reasoning during post-training. Recent work sugges...
arXiv:2607. 00152v1 Announce Type: cross Abstract: Three of the most popular methods for training language models to reason look like three different tricks.
arXiv:2511. 21692v3 Announce Type: replace-cross Abstract: We investigate how well large language models (LLMs) generalize across different task difficulties, a key question for effective data curation and evaluation.
arXiv:2607. 27610v1 Announce Type: new Abstract: Reinforcement learning (RL) finetuning significantly enhances the reasoning capabilities of large language models (LLMs), yet its effectiveness critically depends on selecting prompts of appropriate difficulty for the current policy.
arXiv:2606. 17706v1 Announce Type: cross Abstract: Curriculum learning couples two design choices, how samples are scored by difficulty and how harder samples are paced into training, making it difficult to attribute observed gains to either component.
arXiv:2610.15328v1 Announce Type: cross Abstract: We introduce Anchor-and-Harvest Curriculum (AHC), a domain-informed, rule-based training system for Craftax-Full. AHC combines achievement-oriented t...
arXiv:2606. 07950v1 Announce Type: new Abstract: RL with verifiable rewards can substantially improve LLM reasoning, yet standard GRPO-style training often treats easy, hard, and learnable questions alike through uniform sampling and weighting, leading to inefficient compute allocation.
arXiv:2609.09099v1 Announce Type: new Abstract: Curriculum learning is governed by several coupled design choices---how difficulty is defined, how examples are ordered, how much exposure each level r...
The study clusters 5,000 EdNet-KT3 learners into eight study‑strategy groups based on early‑session behaviors such as resource use, revision, video watching, and problem practice. These clusters predict later engagement metrics—like continued practice and session completion—but do not reliably forecast later unassisted accuracy or mastery. The findings suggest that behavioral clustering captures learning styles and engagement patterns rather than knowledge gains.
arXiv:2608. 20256v1 Announce Type: new Abstract: Reasoning language models trained with reinforcement learning typically operate under a fixed token budget rather than an explicitly adaptive one, which can lead to over-computation on easy problems and insufficient computation on difficult ones.
The paper argues that a benchmark is more than a set of test questions; it defines an empirical distribution over difficulty. It identifies two failure modes—measure mismatch, where the benchmark’s difficulty distribution diverges from the intended one, and coverage gaps, where parts of the difficulty spectrum are missing. The authors propose diagnostic metrics with theoretical guarantees, analyze 25 benchmarks across seven domains, and introduce BenchFill, an algorithm that rewrites items to fill coverage gaps.