41554 blogs · [ { "id": "01a08793-88bd-7122-8f52-b36bca29533b", "title": "Reinforcement Learning for Knowledge Awareness", "url": "https://kalomaze.bearblog.dev/rl-for-knowledge-awareness/", "published_at": "2026-05-07T02:07:00+00:00" }, { "id": "01a08793-88bd-7122-8f52-b36bcac4d87c", "title": "Don't Exclude Rollouts From Your RL Training Runs", "url": "https://kalomaze.bearblog.dev/dont-exclude-rl-rollouts/", "published_at": "2026-02-20T00:37:00+00:00" }, { "id": "01a08793-88bd-7122-8f52-b36bcb2758f8", "title": "RL Learning with LoRA: A Diverse Deep Dive", "url": "https://kalomaze.bearblog.dev/rl-lora-ddd/", "published_at": "2025-11-09T02:53:00+00:00" }, { "id": "01a08793-88bd-7122-8f52-b36bcb3cd52d", "title": "Understanding Transformers... (beyond the Math)", "url": "https://kalomaze.bearblog.dev/understanding-transformers-beyond-the-math/", "published_at": "2025-03-09T00:10:00+00:00" }, { "id": "01a08793-88bd-7122-8f52-b36bcbc9bf72", "title": "GRPO Judge Experiments: Findings & Empirical Observations", "url": "https://kalomaze.bearblog.dev/grpo-judge-experiments-findings-and-empirical-observations/", "published_at": "2025-03-03T22:07:00+00:00" }, { "id": "01a08793-88be-7223-a9a0-eed291351760", "title": "Why does GRPO work?", "url": "https://kalomaze.bearblog.dev/why-does-grpo-work/", "published_at": "2025-02-27T05:45:00+00:00" }, { "id": "01a08793-88be-7223-a9a0-eed2916d1327", "title": "Synthetic rejected preference data creation [via Qwen7b finetune]", "url": "https://kalomaze.bearblog.dev/synthetic-rejected-preference-data-creation-via-qwen7b-finetune/", "published_at": "2025-02-27T05:11:00+00:00" } ] posts Claim your blog

Blog · corpus.blog/blogs/kalomaze.bearblog.dev/posts