Tag: #reinforcement-learning
Writing on GPUs, LLMs, MLOps, Kubernetes — and mindset · 29 posts
29 posts are tagged #reinforcement-learning; the most recent was published on 2026-09-06.
Often tagged together: #ai 23 #deep-learning 21 #ai-papers 4 #dqn 4 #llm 3
Most read in the last 90 days:
- Modern Reinforcement Learning Ecosystems 2026 Deep Dive - RLlib (Anyscale) · Stable-Baselines3 · Tianshou · CleanRL · OpenSpiel (DeepMind) · Gymnasium (Farama) · Acme · PufferLib · Pearl (Meta) · TorchRL
- How Reasoning Models Are Built — From Chain-of-Thought to GRPO, in Twelve Papers
- [Deep RL] 18. AlphaGo Zero: AI That Learns by Playing Itself
How Reasoning Models Are Built — From Chain-of-Thought to GRPO, in Twelve Papers
o1, R1 and QwQ are not new architectures. They are long output optimised toward a gradeable objective. This traces the lineage through twelve papers — Chain-of-Thought (2022), STaR, PRM, DeepSeek-R1's GRPO, s1's budget f
2026-09-06 · 12 min read #ai#llm#reasoning#reinforcement-learning#grpoThe Conditions Under Which a 9B Fine-Tuned for 500 Dollars Beat the Frontier — And How Narrow They Are
On July 28, 2026, a post scored 336 points on Hacker News. It reports that Fermisense trained Qwen3.5-9B with GRPO on roughly 500 dollars worth of GPU time and beat five frontier configurations — using the same tools and
2026-07-31 · 11 min read #ai#llm#fine-tuning#reinforcement-learning#inference-costRobot Safety and Alignment — Trusting Powerful Robots
How can we trust increasingly powerful robots. We take a balanced look at physical safety, constrained reinforcement learning and safety layers, handling distribution shift, human-robot collaboration safety, verification
2026-06-29 · 17 min read #ai-papers#robotics#safety#alignment#reinforcement-learningHow Robots Learn — Imitation Learning and Reinforcement Learning
An overview of the four ways robots acquire skills, followed by a deep look at imitation learning (teleoperation, behavioral cloning, DAgger) and reinforcement learning (rewards, policies, exploration): their principles,
2026-06-29 · 16 min read #ai-papers#robotics#imitation-learning#reinforcement-learning#robot-learningHumanoid Whole-Body Control — Walking on Two Legs and Handling with Two Hands ♪ Listenable
From walking on two legs to handling objects with hands, we lay out the core ideas of bipedal locomotion and whole-body control. We cover ZMP and MPC, reinforcement-learning locomotion, balance and fall recovery, the int
2026-06-29 · 21 min read #ai-papers#robotics#humanoid#locomotion#whole-body-controlCan an Imitation Machine Discover — The Sutton Provocation and the RL Counterattack ♪ Listenable
Rich Sutton, the godfather of reinforcement learning, set the community ablaze in June 2026 by arguing that generative AI trained with supervised learning is an imitation model and therefore limited for genuinely new sci
2026-06-12 · 22 min read #ai#reinforcement-learning#llm#research#ai-agentModern Reinforcement Learning Ecosystems 2026 Deep Dive - RLlib (Anyscale) · Stable-Baselines3 · Tianshou · CleanRL · OpenSpiel (DeepMind) · Gymnasium (Farama) · Acme · PufferLib · Pearl (Meta) · TorchRL ♪ Listenable
A complete deep dive into the reinforcement learning ecosystem as of May 2026. We cover general-purpose RL libraries (RLlib, Stable-Baselines3, Tianshou, CleanRL, TorchRL, PFRL, TF-Agents), the DeepMind stack (OpenSpiel,
2026-05-16 · 21 min read #reinforcement-learning#rllib#stable-baselines-3#tianshou#cleanrl[Deep RL] 20. Deep RL Summary: Algorithm Comparison and Selection Guide
Systematic comparison and selection criteria for deep RL algorithms: value-based, policy-based, Actor-Critic, and model-based approaches
2026-03-19 · 9 min read #reinforcement-learning#deep-learning#ai[Deep RL] 17. Model-Based Reinforcement Learning: Imagination-Augmented Agent
Principles of model-based RL and I2A (Imagination-Augmented Agent): learning environment models and planning through imagination
2026-03-19 · 8 min read #reinforcement-learning#deep-learning#ai[Deep RL] 14. Continuous Action Spaces: DDPG and Distributional Policies
Methods for handling continuous action spaces: A2C extension, DDPG deterministic policy gradient, and distributional policy gradient
2026-03-19 · 7 min read #reinforcement-learning#deep-learning#ai[Deep RL] 11. A3C: Asynchronous Advantage Actor-Critic
How A3C solves the correlation problem of A2C through asynchronous parallel learning: data parallelism and gradient parallelism implementation
2026-03-19 · 7 min read #reinforcement-learning#deep-learning#ai[Deep RL] 08. Building a Stock Trading Agent with Reinforcement Learning
Define stock trading as a reinforcement learning problem, design a custom trading environment, and train trading agents using feedforward and CNN models.
2026-03-19 · 19 min read #reinforcement-learning#deep-learning#ai#finance[Deep RL] 05. Bellman Equation and Value Iteration
Understand the concepts of state value and action value, and apply value iteration and Q-learning to FrozenLake based on the Bellman optimality equation.
2026-03-19 · 9 min read #reinforcement-learning#deep-learning#ai[Deep RL] 02. Getting Started with Reinforcement Learning Using OpenAI Gym
Understand the structure and API of OpenAI Gym, and practice with random agents and wrappers in the CartPole environment.
2026-03-19 · 10 min read #reinforcement-learning#deep-learning#ai[Deep RL] 18. AlphaGo Zero: AI That Learns by Playing Itself
Core principles of AlphaGo Zero: MCTS and self-play, plus implementing a Connect4 bot
2026-03-19 · 9 min read #reinforcement-learning#deep-learning#ai[Deep RL] 15. Trust Region Methods: TRPO, PPO, ACKTR
Trust region techniques for stable policy updates: PPO clipping objective, TRPO constrained optimization, and ACKTR
2026-03-19 · 8 min read #reinforcement-learning#deep-learning#ai[Deep RL] 12. Training Chatbots with Reinforcement Learning
Combining Seq2Seq models with reinforcement learning for chatbot training: from RNN basics to Self-Critical Sequence Training (SCST)
2026-03-19 · 7 min read #reinforcement-learning#deep-learning#ai#nlp[Deep RL] 09. Policy Gradient: Policy-Based Reinforcement Learning
Understand the difference between value-based and policy-based methods, and apply the mathematical derivation of Policy Gradient and the REINFORCE algorithm to CartPole and Pong.
2026-03-19 · 6 min read #reinforcement-learning#deep-learning#ai#policy-gradient[Deep RL] 06. Deep Q-Network: Principles and Implementation of DQN
Understand the core ideas of DQN (experience replay, target network) that overcome the limitations of table-based Q-learning, and implement it in the Pong environment.
2026-03-19 · 10 min read #reinforcement-learning#deep-learning#ai#dqn[Deep RL] 03. PyTorch Deep Learning Basics: From Tensors to Neural Networks ♪ Listenable
Learn PyTorch tensor operations, automatic differentiation, and neural network components, then implement TensorBoard monitoring and a GAN example using Atari images.
2026-03-19 · 20 min read #reinforcement-learning#deep-learning#ai#pytorch