---
title: "#reinforcement-learning articles | Hashnode"
description: "Explore 262 articles tagged #reinforcement-learning on Hashnode. Tutorials, guides, and how-tos from the community."
url: "https://hashnode.com/tag/reinforcement-learning"
---

Tag

# #reinforcement-learning

262 posts · 54 followers

[Viswaretas Kotra](https://hashnode.com/@viswa-kotra) · [Nov 9](https://hashnode.com/posts/six-wrong-normalization-formulas-passed-a-correctness-check/6af1d05036a54cca5df5fe55)

### [Six wrong normalization formulas passed a correctness check](https://nodus-compute.hashnode.dev/six-wrong-normalization-formulas-passed-a-correctness-check?utm_source=hashnode&utm_medium=feed)

Six incorrect normalization formulas passed with identity inputs and failed with varied parameters in separate synthetic CPU and T4 CUDA runs.

9 min read · [nodus-compute.hashnode.dev](https://nodus-compute.hashnode.dev/six-wrong-normalization-formulas-passed-a-correctness-check?utm_source=hashnode&utm_medium=feed)

[Viswaretas Kotra](https://hashnode.com/@viswa-kotra) · [Nov 9](https://hashnode.com/posts/the-reward-is-part-of-the-dataset/6af192e0befea3fedc931883)

### [The reward is part of the dataset](https://nodus-compute.hashnode.dev/the-reward-is-part-of-the-dataset?utm_source=hashnode&utm_medium=feed)

Six wrong normalization formulas passed a CPU checker with identity inputs. What this means for kernel RL feedback and held-out evaluation.

3 min read · [nodus-compute.hashnode.dev](https://nodus-compute.hashnode.dev/the-reward-is-part-of-the-dataset?utm_source=hashnode&utm_medium=feed)

[Ishan Nerlekar](https://hashnode.com/@ishan-) · [Nov 7](https://hashnode.com/posts/the-verifier-is-the-weakest-link-prime-intellect-s-rl-environments/6aef06eaeddf27744a4d09b2)

### [The verifier is the weakest link: auditing Prime Intellect's RL environments](https://thegradersaidyes.hashnode.dev/the-verifier-is-the-weakest-link-prime-intellect-s-rl-environments?utm_source=hashnode&utm_medium=feed)

Verifier auditing for Prime Intellect

10 min read · [thegradersaidyes.hashnode.dev](https://thegradersaidyes.hashnode.dev/the-verifier-is-the-weakest-link-prime-intellect-s-rl-environments?utm_source=hashnode&utm_medium=feed)

[ym p](https://hashnode.com/@hurt) · [Oct 22](https://hashnode.com/posts/a-successful-code-repair-can-still-have-a-negative-advantage/6ad99c8d217fe726b407abe7)

### [A Successful Code Repair Can Still Have a Negative Advantage](https://ai-coding.hashnode.dev/a-successful-code-repair-can-still-have-a-negative-advantage?utm_source=hashnode&utm_medium=feed)

Debug RLEF advantages with Python: separate rewards, returns, and value baselines, and see why a successful code repair can have a negative advantage.

6 min read · [ai-coding.hashnode.dev](https://ai-coding.hashnode.dev/a-successful-code-repair-can-still-have-a-negative-advantage?utm_source=hashnode&utm_medium=feed)

[Yunus Emre Vurgun](https://hashnode.com/@yunusemrevurgun) · [Oct 18](https://hashnode.com/posts/your-256-byte-q-table-is-lying-to-you-rounding-not-memory-is-what-breaks-rl-on-microcontrollers/6ad476567bd3311572d5b7a0)

### [Your 256-byte Q-table is lying to you: rounding, not memory, is what breaks RL on microcontrollers](https://yunusemrevurgun.hashnode.dev/your-256-byte-q-table-is-lying-to-you-rounding-not-memory-is-what-breaks-rl-on-microcontrollers?utm_source=hashnode&utm_medium=feed)

A reproducible gridworld experiment on float32, int8, and int16 Q-tables, and the one-line fix that rescues int8

13 min read · [yunusemrevurgun.hashnode.dev](https://yunusemrevurgun.hashnode.dev/your-256-byte-q-table-is-lying-to-you-rounding-not-memory-is-what-breaks-rl-on-microcontrollers?utm_source=hashnode&utm_medium=feed)

[Dr. Juan Zamora-Mora](https://hashnode.com/@doczamora) · [Oct 12](https://hashnode.com/posts/reinforcement-learning-lunar-lander-with-dqn/6acd2b482af13e75f970e622)

### [Reinforcement Learning - Lunar Lander with DQN](https://www.doczamora.com/reinforcement-learning-lunar-lander-with-dqn?utm_source=hashnode&utm_medium=feed)

This is a classic control problem with no rules written by hand. The task: land a spacecraft gently on a landing pad on the Moon.

16 min read · [doczamora.com](https://www.doczamora.com/reinforcement-learning-lunar-lander-with-dqn?utm_source=hashnode&utm_medium=feed)

[Aleksei Romanov](https://hashnode.com/@gfactor) · [Oct 9](https://hashnode.com/posts/multi-reward-rl-part-3-gdpo-cispo-repo-r-at-27b-and-the-advantage-floor-that-stopped-learning/6ac929e55952e02b451b58c9)

### [Multi-Reward RL, Part 3: GDPO + CISPO + REPO-R at 27B, and the Advantage Floor That Stopped Learning](https://gfactor.hashnode.dev/multi-reward-rl-part-3-gdpo-cispo-repo-r-at-27b-and-the-advantage-floor-that-stopped-learning?utm_source=hashnode&utm_medium=feed)

Qwen3.8-27B, 600 steps, 12 reward channels: CISPO, REPO-R and prompt filtering lifted holdout from −0.47 to +1.80, and an advantage floor silently zeroed 96% of the signal.

15 min read · [gfactor.hashnode.dev](https://gfactor.hashnode.dev/multi-reward-rl-part-3-gdpo-cispo-repo-r-at-27b-and-the-advantage-floor-that-stopped-learning?utm_source=hashnode&utm_medium=feed)

[Nick One-Drop](https://hashnode.com/@nickonedrop) · [Oct 7](https://hashnode.com/posts/multi-agent-cooperation-through-in-context-co-player-inference-2026/6ac684f0d573c78a39c071fe)

### [Multi-agent cooperation through in-context co-player inference (2026)](https://nickonedrop.hashnode.dev/multi-agent-cooperation-through-in-context-co-player-inference-2026?utm_source=hashnode&utm_medium=feed)

This paper takes a decentralized multi-agent reinforcement learning approach. What does that mean here? For context, let’s look at another DeepMind paper with overlapping authors: Multi-agent cooperat

3 min read · [nickonedrop.hashnode.dev](https://nickonedrop.hashnode.dev/multi-agent-cooperation-through-in-context-co-player-inference-2026?utm_source=hashnode&utm_medium=feed)

[Aleksei Romanov](https://hashnode.com/@gfactor) · [Oct 6](https://hashnode.com/posts/multi-reward-rl-part-2-benchmarking-grpo-dapo-and-cispo-on-unseen-tasks/6ac5597f65f0cea9c5cc433d)

### [Multi-Reward RL, Part 2: Benchmarking GRPO, DAPO, and CISPO on Unseen Tasks](https://gfactor.hashnode.dev/multi-reward-rl-part-2-benchmarking-grpo-dapo-and-cispo-on-unseen-tasks?utm_source=hashnode&utm_medium=feed)

Qwen3-14B DEX experiments compare GRPO, DAPO, GDPO, CISPO, ADAPO and REPO-R with and without thinking: full holdout results, reward curves, time and memory.

14 min read · [gfactor.hashnode.dev](https://gfactor.hashnode.dev/multi-reward-rl-part-2-benchmarking-grpo-dapo-and-cispo-on-unseen-tasks?utm_source=hashnode&utm_medium=feed)

[Aleksei Romanov](https://hashnode.com/@gfactor) · [Oct 5](https://hashnode.com/posts/the-verifier-design-playbook-how-to-build-rlvr-gyms-that-models-can-t-game/6ac424015ee1341e55bcf41c)

### [The Verifier Design Playbook: How to Build RLVR Gyms That Models Can't Game](https://gfactor.hashnode.dev/the-verifier-design-playbook-how-to-build-rlvr-gyms-that-models-can-t-game?utm_source=hashnode&utm_medium=feed)

In RLVR the verifier is your loss function. How to build deterministic gyms: 3-layer design, mutation checks, cache isolation, and reward-hacking defenses.

14 min read · [gfactor.hashnode.dev](https://gfactor.hashnode.dev/the-verifier-design-playbook-how-to-build-rlvr-gyms-that-models-can-t-game?utm_source=hashnode&utm_medium=feed)
