Ashutosh Kumar Sinha

New AI methods boost model robustness in shifting environments

Researchers have developed new methods for unsupervised domain adaptation (UDA) to improve the robustness of AI models in dynamic environments. One approach, SFT+RL, uses supervised fine-tuning and reinforcement learnin…