arXivMartin Marek, Dongkyu Cho, Shikai Qiu, Rumi Chunara, Pavel Izmailov, Andrew Gordon WilsonMon, May 25, 2026, 10:54 AM PDT
score 16.7
597likes91RT16reply
Language models prevent forgetting old skills using self-generated data
Original: Forgetting in Language Models: Capacity, Optimization, and Self-Generated Replay
Source: arxiv.org ↗
Writing ELI5 summary…