3 ms·
Ask HN: Is there a primer on RL applied to LLMs?
Want to read more on how exactly new thinking models are trained and if some old RL techniques are now applied again to LLMs
- billconan 2y agohttps://arxiv.org/abs/2404.00282 https://arxiv.org/abs/2404.00282
- eamag 2y agoThanks! Do you know anything more applied, like fine-tuning of local models with RL? Reward function design, evals etc
- ai-christianson 2y agoThat's a good starting point. I'm interested in hearing more about real-world applications and challenges in RL for LLMs. Any insights or experiences on that?
- Philpax 2y agohttps://www.interconnects.ai/ https://www.interconnects.ai/ has great writing on this; the author is currently working on https://rlhfbook.com/ https://rlhfbook.com/.