curriculum learning
Curriculum learning is an approach in machine learning where models are trained on increasingly complex tasks or data distributions. This structured learning process can enhance the learning efficiency and performance of AI systems.
- AceReason-Nemotron: Advancing Math and Code Reasoning through Reinforcement Learning
- Angles Don’t Lie: Unlocking Training‑Efficient RL Through the Model’s Own Signals
- Bandit Guided Submodular Curriculum for Adaptive Subset Selection
- Counterfactual Evolution of Multimodal Datasets via Visual Programming
- Curriculum Design for Trajectory-Constrained Agent: Compressing Chain-of-Thought Tokens in LLMs
- D-VST: Diffusion Transformer for Pathology-Correct Tone-Controllable Cross-Dye Virtual Staining of Whole Slide Images
- DeepDiver: Adaptive Web-Search Intensity Scaling via Reinforcement Learning
- DeepKD: A Deeply Decoupled and Denoised Knowledge Distillation Trainer
- Hierachical Balance Packing: Towards Efficient Supervised Fine-tuning for Long-Context LLM
- Learning to Solve Complex Problems via Dataset Decomposition
- Mean Flows for One-step Generative Modeling
- NAVIX: Scaling MiniGrid Environments with JAX
- On the creation of narrow AI: hierarchy and nonlocality of neural network skills
- SATURN: SAT-based Reinforcement Learning to Unleash LLMs Reasoning
- Shaping Sequence Attractor Schema in Recurrent Neural Networks
- Towards A Generalist Code Embedding Model Based On Massive Data Synthesis
- Transformers Provably Learn Chain-of-Thought Reasoning with Length Generalization
- When Does Curriculum Learning Help? A Theoretical Perspective
- World-aware Planning Narratives Enhance Large Vision-Language Model Planner