2017 · Reinforcement learning
Proximal Policy Optimization Algorithms
Proximal Policy Optimization proposes a policy-gradient objective that limits overly large updates while simplifying implementation.
Paper & contextJohn Schulman
Reinforcement learning
Improving policies while controlling the size of each update.
Schulman studies reinforcement learning and how to train useful language-model behavior. This selection links trust-region and proximal policy optimization with human-feedback training, reward-model limitations, and evaluation of model responses.
Selected work
10 papers
2015 · Reinforcement learning
Trust Region Policy Optimization
Selected research in reinforcement learning, robotics. Read the full paper, including the methods, experiments, and reported results.
Paper & context2022 · Alignment
Training language models to follow instructions with human feedback
InstructGPT studies fine-tuning language models using demonstrations and human preference judgments to better follow instructions.
Paper & context2016 · Reinforcement learning
OpenAI Gym
Selected research in reinforcement learning. Read the full paper, including the methods, experiments, and reported results.
Paper & context2022 · Reinforcement learning
Scaling Laws for Reward Model Overoptimization
Selected research in reinforcement learning. Read the full paper, including the methods, experiments, and reported results.
Paper & context2023 · LLMs
GPT-4 Technical Report
The GPT-4 technical report documents capabilities, evaluations and limitations of a multimodal model, while withholding many architecture and training details.
Paper & context2026 · Reinforcement learning
Chunky Post-Training: Data Driven Failures of Generalization
Selected research in reinforcement learning. Read the full paper, including the methods, experiments, and reported results.
Paper & context2025 · Reinforcement learning
Detecting Adversarial Fine-tuning with Auditing Agents
Selected research in reinforcement learning. Read the full paper, including the methods, experiments, and reported results.
Paper & context2025 · Reinforcement learning
Stress-Testing Model Specs Reveals Character Differences among Language Models
Selected research in reinforcement learning. Read the full paper, including the methods, experiments, and reported results.
Paper & context2025 · Scaling laws
Reasoning Models Don't Always Say What They Think
Selected research in scaling laws, reinforcement learning. Read the full paper, including the methods, experiments, and reported results.
Paper & contextAn independent editorial profile. Inclusion does not imply Council membership or endorsement. Research is collaborative; coauthorship does not imply sole credit.
Selection & sources