AutoScaffold: AI-Guided Scaffolding for Reinforcement Learning of Language Models
cs.LG, cs.CL
Submitted: 2026-11-10
Updated: 2026-11-10
Code: https://github.com/karpathy/autoresearch
Terminology
Sources
- Back to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMs
- From Trainee to Trainer: LLM-Designed Training Environment for RL with Multi-Agent Reasoning
- EvoTrainer: Co-Evolving LLM Policies and Training Harnesses for Autonomous Agentic Reinforcement Learning
- LLMZero: Discovering Adaptive Training Strategies for RL Post-Training via LLM Agents
- ReSkill: Reconciling Skill Creation with Policy Optimization in Agentic RL
- Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Steps
- Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning
- TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension
- Privileged Information Dropout in Reinforcement Learning
- Meta-Harness: End-to-End Optimization of Model Harnesses
- QuestA: Expanding Reasoning Capacity in LLMs via Question Augmentation
- StitchCUDA: An Automated Multi-Agents End-to-End GPU Programing Framework with Rubric-based Agentic Reinforcement Learning
- Self-Hinting Language Models Enhance Reinforcement Learning
- SKILLC: Learning Autonomous Skill Internalization in LLM Agents via Contrastive Credit Assignment
- Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses
- Large Language Model Agent for Hyper-Parameter Optimization
- Dr. Kernel: Reinforcement Learning Done Right for Triton Kernel Generations
- Understanding R1-Zero-Like Training: A Critical Perspective
- The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery
- SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis