Tag: Reinforcement Learning

All the articles with the tag "Reinforcement Learning".

• From Long CoT to Agent Swarms: The Documented Evolution of Kimi's Reinforcement Learning

  18 Jul, 2026 · 15 min read

  A source-grounded history of Kimi's reinforcement-learning stack, from Kimi k1.5's long-context outcome RL and partial rollouts to K2's general RL and K2.5's multimodal GRMs and Parallel-Agent RL.

• Reproducing CompactionRL: From the GLM-5.2 Algorithm to a Live slime E2E

  17 Jul, 2026 · 13 min read

  A source-grounded reproduction of CompactionRL: the algorithm, missing recipe details, a 96-step multi-seed causal experiment, and a live Qwen actor-critic update through THUDM/slime on four A10 GPUs.

• Training the Critic Without Crashing the Reward: A Practical Guide to Agentic RL

  12 Jul, 2026 · 20 min read

  A practical framework for critic training and credit assignment in long-horizon LLM agents: IQL, pairwise advantage, hindsight and counterfactual critics, privileged information, turn-level MDPs, chain-of-thought monitoring, and reward-crash diagnosis.

• From GRPO Outcome Rewards to Token-Level Advantage

  8 Jul, 2026 · 20 min read

  A practical framework for turning GRPO-style sequence rewards into token-level advantages, including GAE-style estimators, credit assignment routes, and multi-reward training design.

• Scaling RL for White-Collar Work: The Environment Foundry

  3 Jul, 2026 · 20 min read

  A practical framework for turning common white-collar workflows into RL environments: spreadsheets, CRM tasks, customer support, web research, dashboards, and other software-mediated work.

• The Unverifiable Reward Problem: The Real Frontier of RL for LLMs

  7 Mar, 2026 · 11 min read

  Deep research on tasks with unverifiable rewards in RL — the key bottleneck for scaling RL beyond math and code. Covers JEPO, NRT, RLNVR, self-play methods, GenRM, Constitutional AI, reward hacking mitigation, and more.