2 papers
cs.AI2026
Scaling In-Context Online Learning Capability of LLMs via Cross-Episode Meta-RL
Xiaofeng Lin, Sirou Zhu, Yilei Chen +6
Large language models (LLMs) achieve strong performance when all task-relevant information is available upfront, as in static prediction and instruction-following problems. However…
cs.LG2025
Debunk the Myth of SFT Generalization
Xiaofeng Lin, Hejian Sang, Zhipeng Wang +1
A prevailing view holds that supervised fine-tuning (SFT) memorizes training data and fails to generalize, whereas reinforcement learning (RL) attains broader robustness. We revisi…