Learning to Solve Hard Problems in RL for LLMs by Never Giving Up Sep 15, 2026 Table of Contents What is your eval actually measuring? What causes the Matthew Effect? Never Give Up on hard problems Async RL staleness and tricks for NGU NGU at a bigger scale: Math NGU on a different scale: Code The Matthew Effect is a Primacy Bias, sort of Limitations Conclusion Acknowledgements Citation This is a blog post for my recent paper on RL post-training of LLMs: introducing the Matthew Effect and proposing to solve it with Never Give Up. It is presented interactively and less formally, more like how I give the talk. For a deeper, more technical dive, check out the paper on arxiv and code on github . What is your eval actually measuring? # Every good RL practitioner has no doubt seen an eval curve go up. Here is the AIME 2025 eval during our RL training of Olmo 3.1 RL-Zero Math (1) (1) see Olmo 3.1 blog post and arxiv Training Olmo 3 7B base with RL on Dolci RL-Zero math improves its overall math ability. Or does it?…