1Write programs you can still hack when you feel dumb (opens in new tab)(draketo.de)1xhevahir16d ago0
2Prism: Demystifying Retention and Interaction in Mid-Training (opens in new tab)(arxiv.org)1xhevahir2mo ago0
3Toward Training Superintelligent Software Agents Through Self-Play SWE-RL (Meta) (opens in new tab)(arxiv.org)1xhevahir5mo ago0
